new evaluation of access to primary healthcare · 2010. 4. 14. · jeny shrestha february, 2010 ....

Evaluation of Access to Primary Healthcare

A Case Study of Yogyakarta, Indonesia

Jeny Shrestha

February, 2010

Evaluation of Access to Primary Healthcare

A Case Study of Yogyakarta, Indonesia

by

Jeny Shrestha

Thesis submitted to the International Institute for Geo-information Science and Earth Observation in

partial fulfilment of the requirements for the degree of Master of Science in Geo-information Science

and Earth Observation, Specialisation: (Urban Planning and Management)

Thesis Assessment Board

Chairman : Prof. Dr. Ir.M.F.A.M. van Meerseveen

External examiner : Prof. Dr. O. Verkoren

First Supervisor : Dr. S. Amer

Second Supervisor : Dr. J.A. Martinez

INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION

ENSCHEDE, THE NETHERLANDS

Disclaimer

This document describes work undertaken as part of a programme of study at the International

Institute for Geo-information Science and Earth Observation. All views and opinions expressed

therein remain the sole responsibility of the author, and do not necessarily represent those of the

institute.

i

Abstract

Access to primary healthcare (PHC) has been a major goal of much health legislation and planning

processes in order to meet the healthcare need of population. Improvements in access to PHC pave the

way for advancement in the quality of people‟s life. However, efforts to conceptualize and

operationalize measures of access have varied. In this study, access is decomposed into five

dimensions as: availability, accessibility, affordability, acceptability and adequacy. One of the

motivations for this concept is the presence of relatively limited studies in access to healthcare

considering non-spatial factors like acceptability and adequacy. The relevance of access in this study

is substantive, in investigating for each of these dimensions in how far the distribution in space and

people is equitable.

Using primary data collected by household survey, variation in dimensions of access was measured,

by developing objective and subjective indicators, in three villages in the Province of Yogyakarta

(DIY). Variation of access was evaluated across different health facilities users and socioeconomic

classes using descriptive and explorative statistics. Existing situation of access to PHC was compared

with the related health policies in DIY. Using official census data of 2005, a comparative analysis

with primary data was done to complement the disaggregated data with aggregated one.

The results of this study showed that variation in access exists between villages and across different

socioeconomic class. Physical accessibility and affordability in general was not very problematic as

the result of effective policy implementation. In general people had no real preference over cultural

factors and gender of medical staffs (under acceptability). The analyses demonstrate that waiting time

in health facilities (under availability) and inter-personal treatment from medical staffs (under

adequacy) were important causes of dissatisfaction with access. These issues require policy attention

to further improve access to PHC. While developing health policies, consideration to internationally

accepted health standards can be effective in meeting minimum required standard for access to

healthcare. Results of analysis in scale effect revealed that aggregated census data tends to average

out variations as compared to data obtained at higher spatial resolution. Hence, care should be taken

before drawing conclusions.

In general, findings of this study indicate how dimensions in access can be quantified and measured

for the evaluation of access to PHC across different population groups. Also the possible effect of

scale in results of analyses is demonstrated. The approaches and findings of this study can be useful in

addressing problematic issues in access to PHC in the region.

Keywords: Primary healthcare, Access, Dimensions of access, Socio-economic stratification, Scale

effect

ii

Acknowledgements

This research is the result of valuable support from many institutions and individuals. Their

contributions in different ways have helped in the successful completion of this research.

I am very thankful to God who gives me strength and blessings every day of my life. My family

deserves special thanks for their love, support and faith in me besides the inspiration that they have

provided. I would like to thank Netherlands Fellowship Programme for the opportunity to pursue MSc

study in ITC which has broadened my academic knowledge and professional skills.

I would like to express my sincere gratitude to my supervisors, Dr. S. Amer and Dr. J.A. Martinez for

their guidance and critical suggestions from the inception till completion of this thesis. Their valuable

suggestions at various stages of this research were crucial for conducting my work in the right

direction. Their academic and moral support has been very inspiring throughout this study. I am also

thankful to all UPM staffs and teachers for the excellent academic guidance and assistance during the

programme. Special thanks to Ir. Mark Brussel for initial inspiration and introducing us to the city of

Yogayakarta. His guidance during fieldwork was very helpful.

I am very much thankful to „Pustral‟, Center for Transportation and Logistics Studies, Gadjah Mada

University for providing me necessary data and information, which were of immense value in this

study. Special thanks to my thesis advisor, Ir. Arif Wismadi for his suggestions and all administrative

support during fieldwork. I must thank Dr. Choirul Anwar, the Chief of Health Department,

Yogyakarta for his valuable time in providing me relevant information and documents. My

acknowledgement goes to Dimas, Pugo and Tamzil for their hard work in conducting field surveys

and also to all participants of the interviews. The hospitality shown by the city of Yogya and

especially all staffs of Pustral is highly appreciated.

My thanks go to all my colleagues and fellow Nepalese friends; Arun, Gopi, Diwakar, GRD, Ganesh,

Janak and Jay for their help, cooperation and good memories shared together during this course of

study. Special thanks to my friend Jiwan Limbu for encouragement throughout. All wonderful time

shared with friends and colleagues were valuable for providing an extremely diverse, interactive, fun-

filled and an intellectual rendezvous.

iii

Table of contents

1. Introduction ................................................................................................................................ 1

1.1. Background of Study .......................................................................................................... 1

1.2. Research Problem ............................................................................................................... 2

1.3. Research Objectives ........................................................................................................... 3

1.4. Research Questions ............................................................................................................. 3

1.5. Research Framework .......................................................................................................... 3

1.6. Structure of Report ............................................................................................................. 4

2. Review on Access to Primary Healthcare ................................................................................... 6

2.1 Introduction ........................................................................................................................ 6

2.2.1 Definitions and Concepts of Access to Primary Healthcare ......................................... 6

2.2.2 Dimensions of Access to Healthcare Service ............................................................... 7

2.3 Conceptual Framework for Access to Healthcare .......................................................... 11

2.3.1 Healthcare Utilization and Quality of Care ................................................................... 12

2.3.2 Equity in Access to Primary Heathcare ......................................................................... 13

2.4 Measuring Dimensions of Access to Primary Healthcare .................................................. 14

2.4.1 Developing Indicators to Quantify and Measure Dimensions of Access ..................... 14

2.4.2 Subjective and Objective Indicators .......................................................................... 15

2.4.3 Analyzing and Measuring Indicators ......................................................................... 16

2.5 Methodological Problems to be Resolved in Evaluating Access ....................................... 17

2.5.1 Effect of Scale in Analysis ........................................................................................ 18

2.6 Conclusion........................................................................................................................ 19

3. Study Area Description and Healthcare Policies ...................................................................... 21

3.1 General Description of Study Area ................................................................................... 21

3.1.1 Demographic Condition ................................................................................................ 21

3.1.2 Landuse and Economic Activities ................................................................................. 23

3.1.3 Administrative Units ..................................................................................................... 23

3.2 Health Policies and Planning Systems .............................................................................. 24

3.2.1 Decentralization ........................................................................................................... 24

3.2.2 Health Policies and Strategies ...................................................................................... 25

3.2.3 Health Organization System ......................................................................................... 27

3.2.4 Planning Process for Healthcare Service ....................................................................... 28

4. Research Methodology ............................................................................................................. 29

4.1 Research Design ............................................................................................................... 29

4.2 Fieldwork Preparation ...................................................................................................... 29

4.2.1 Study Area Selection ................................................................................................ 31

4.2.2 Sampling Strategy ..................................................................................................... 32

4.3 Field Work ....................................................................................................................... 32

4.3.1 Primary Data............................................................................................................. 32

4.3.2 Secondary Data ......................................................................................................... 35

4.4 Post Fieldwork ................................................................................................................. 35

4.5 Challenges During Fieldwork ........................................................................................... 35

4.6 Data Analysis ................................................................................................................... 36

iv

4.6.1 Socioeconomic Stratification of Sample Households ................................................. 36

4.6.2 Measuring Dimensions of Access to PHC .................................................................. 37

4.6.3 Synthesising Indicators .............................................................................................. 37

4.6.4 Access to PHC in Relation to Existing Health Policies .............................................. 38

4.6.5 Scale Effect in Analyzing Socioeconomic and Access Variation ............................... 38

5. Perceived Access to PHC at Micro and Macro Level ............................................................... 40

5.1 Household Characteristics and Socioeconomic Stratification ............................................ 40

5.2 Measuring Dimensions of Access to PHC .......................................................................... 45

5.2.1 Descriptive Statistics for Availability of PHC ........................................................... 45

5.2.2 Descriptive Statistics for Accessibility to PHC .......................................................... 47

5.2.3 Descriptive Statistics for Affordability of PHC .......................................................... 47

5.2.4 Descriptive Statistics for Acceptability and Adequacy of PHC .................................. 48

5.2.5 Overall Satisfaction Level with Access to PHC ......................................................... 50

5.2.6 Influencing Factors to Overall Satisfaction Level ...................................................... 51

5.3 Synthesis of Indicators to Develop Summary Scores .......................................................... 53

5.4 Existing Situation of Access to PHC from Policy Perspective ........................................... 58

5.5 Scale Effect in Analyzing Socioeconomic Attributes and Access to PHC.......................... 61

5.5.1 Comparing Individual Variables from Census and Primary data ................................ 61

5.5.2 Comparing Variation within Village with Variation at Sub district Level .................. 64

6. Discussions on Findings ............................................................................................................ 66

6.1 Sub-objective 1: Measuring Access to PHC at Micro Level .............................................. 66

6.2 Sub-objective 2: Existing State of Access in Relation to Health Policy ............................. 68

6.3 Sub-objective 3: Variations in Results Obtained from Census and Primary Data ............... 69

7. Conclusions and Recommendations .......................................................................................... 70

7.1 Conclusions ....................................................................................................................... 70

7.1.1 Main Findings from Sub-objective 1 .......................................................................... 70



7.2 Recommendations ............................................................................................................. 72

References ........................................................................................................................................ 73

Appendix A: Empirical references to indicators used in this study .................................................... 76

Appendix B: Description and rationales of indicators used in this study ........................................... 79

Appendix C: Content of household survey questionnaire .................................................................. 83

Appendix D: Content of Interview with Health Facility Personnel .................................................... 88

Appendix E: Distribution of summary scores of dimensions of access .............................................. 90

v

List of figures

Figure 1-1: Conceptual Framework .................................................................................................. 4

Figure 2-1: The health access framework [Source: Obrist, Iteba et al.(2007)] .............................. 11

Figure 2-2: Conceptual framework to evaluate access to PHC ...................................................... 12

Figure 3-1: Population growth trend in DIY (1971 – 2000) ........................................................... 22

Figure 3-2: Population density map of DIY .................................................................................... 22

Figure 3-3: Land use of DIY............................................................................................................ 24

Figure 3-4: Administrative boundary.............................................................................................. 24

Figure 3-5: Organizational Structure of Health System in Indonesia............................................. 27

Figure 4-1: Location of surveyed household in villages ................................................................. 33

Figure 4-2: Data collection and fieldwork observations ................................................................ 34

Figure 5-1: Proportion of socioeconomic cluster per village ......................................................... 43

Figure 5-2: General characteristics of socioeconomic clusters per village .................................... 44

Figure 5-3: Spatial distribution of socioeconomic classes ............................................................. 44

Figure 5-4: Percentage of subjective perception on waiting time in PHC ...................................... 46

Figure 5-5: Percentage of subjective perception on travel distance and travel time to PHC ......... 47

Figure 5-6: Percentage of subjective perception on total cost for PHC ......................................... 48

Figure 5-7: Percentage of subjective perception on factors related to acceptability and adequacy

........................................................................................................................................................ 50

Figure 5-8: Perceived satisfaction level with access to PHC .......................................................... 51

Figure 5-9: Synthesising indicators to develop summary scores for dimensions of access to PHC 54

Figure 5-10: Summary score chart for dimensions of access to PHC ............................................ 56

Figure 5-11: Standardized residuals for perceived satisfaction with various factors of access ..... 57

Figure 5-12: Geographic distance to health facilities per village in DIY, ...................................... 60

Figure 5-13: Population-doctor ratio per sub districts in DIY ....................................................... 60

Figure 5-14: Euclidean distance from households to heath facilities ............................................. 63

vi

List of tables

Table 2-1 : Empirical references to dimensions of access ............................................................... 10

Table 3-1: Population density per regency in DIY from years 2003 – 2006 .................................... 22

Table 3-2: General demographic characteristics of DIY ................................................................. 22

Table 3-3: National standards for health service, Indonesia .......................................................... 26

Table 4-1: Research Design............................................................................................................. 31

Table 4-2: Collected data from secondary sources ......................................................................... 36

Table 5-1: Statistics of different socioeconomic clusters................................................................. 42

Table 5-2: Descriptive statistics of waiting time in health facilities ................................................ 46

Table 5-3: Correlation matrices between perception on factors of access and overall satisfaction

level with access to PHC .................................................................................................................. 53

Table 5-4: Summary score for dimensions of access per socioeconomic class ................................ 58

Table 5-5: Common variables in census 2005 and household data ................................................ 62

Table 5-6: Distance to healthcare facilities from Census and primary data ................................... 64

Table 5-7: Socioeconomic and access to PHC variability in terms of CV ....................................... 65

vii

Acronyms

DIY Province of Yogyakarta

BPS „Badan Pusat Statistik’ Statistics of Indonesia

IDR Indonesian Rupiah

UGM Gadjah Mada University

PHC Primary healthcare

HSEC High socioeconomic class

MSEC Middle socioeconomic class

LSEC Lower socioeconomic class

GIS Geographic information system

WHO World Health Organization

EVALUATION OF ACCESS TO PRIMARY HEALTHCARE, CASE STUDY IN YOGYAKARTA

1

1. Introduction

This study develops a methodology that contributes in the evaluation of access to primary

healthcare in the province of Yogyakarta (DIY), Indonesia. Study emphasizes on appraisal of

existing healthcare service, to see if variation exists in different dimensions of access. If the

variation is significantly high, then issues of disparity or inequality arise. This research aims to

highlight various dimensions influencing access, and in defining inequalities that arise from such

variation. Consistency in results obtained from different scale of analysis is also checked for the

study area.

1.1. Background of Study

Primary healthcare (PHC) service is an important concern for increasingly growing population

primarily in developing countries. Cities in the developing countries are experiencing unprecedented

growth for decades. The emergent population is aligned with the increasing demand for healthcare

service along with all sorts of other infrastructures and public service provision required to ensure the

basic quality of life. As the result, the situation might come when the existing infrastructure and

public facility will not be able to provide adequate service to the constantly expanding population. “If

population grows faster either through natural growth or movement in areas facilitated with more

services than others, it is possible that the per-capita measure of facility availability could be

worsened” (Yamauchi, Chowdhury et al. 2007, p.19). If continued, this process might gradually lead

towards significant variation further fuelling inequality in service provision.

Adequate level of access to PHC is a major health development issue. Improvements in PHC services

pave the way for advancement in the quality of people‟s life. Yamauchi, Chowdhury et al. (2007)

states that planners have potential role in providing better coordination in the service provision of

such facilities. Equal provision of basic public services like primary healthcare has been a matter of

interest for researchers, planers and policy makers since decades. “The achievement of equity in the

distribution of urban public facilities is a goal of paramount importance to urban planners, who must

analyze whether and to what degree their distribution is equitable” (Tsou, Hung et al. 2005, p.424).

Inequality in healthcare refers to disparities in access to health facility to large extend.

Access to PHC can be seen from a wide range of angles considering different factors and dimensions.

These factors depend largely on the objective of study and the context of study area. Number of

literatures (Penchansky and Thomas 1981; Millman 1993; Guagliardo 2004) have defined such

factors and dimensions as barriers to access while defining and conceptualizing access to healthcare,

which is explained in detail in next chapter. Factors like physical distance, travel time to reach the

facility, availability of transportation, service cost, waiting time, language etc are referred as barriers

because these factors impede people in certain socioeconomic condition to reach and achieve the

healthcare service.

Dimensions in access can be measured by developing indicators, both objective and subjective

indicators. Objective indicators are the observable facts and figures like distance, time, cost etc and

subjective indicators are normally derived from peoples‟ perception and satisfaction level on each

factors obtained from ground survey.

2

Scale of analysis should be considered carefully while analyzing variations in different dimensions of

access to PHC, as results of similar analysis may vary based on the selected areal unit. As spatially

aggregated data obtained from census are often used by researchers and policy makers to analyze

different contextual determinants of health, it is important to understand what factors and criteria

were used to aggregate such data. Aggregation of data might results in the loss of information which

might be relevant in specific study like healthcare. The limitations of scale problems are discussed

later in chapter two based on some empirical studies. This study analyzes and compares the results

from disaggregated primary data to the aggregated secondary data related to socioeconomic and

access to PHC in the study area.

In this study, influencing factors in access to PHC is measured in the Province of Yogyakarta,

Indonesia. Rapid urbanization in Indonesia is compounding the pressures for higher supply of

healthcare services, so is the case in the Province of Yogyakarta having four rural regencies and one

urban, the city of Yogyakarta. Regencies are further divided into sub-district and village being the

smallest administrative boundary. Depending on the physical and socioeconomic condition of people,

relevance of various factors is explored to evaluate the access to PHC. Results obtained from

aggregated and disaggregated data are compared to see the consistency in different scale of data.

1.2. Research Problem

As the result of rapid population growth and socioeconomic heterogeneity, variation in the provision

of PHC services can be assumed to occur. This affects the life of people in lower socioeconomic

group more adversely if the variation is significantly high. It is important to identify and address such

variations in order to provide adequate service to all people regardless of their socio-economic status.

Many studies have been done focusing on physical accessibility to define spatial equity in service

provision. However, there are other factors along with accessibility, which determines equity in

access to service. As mentioned earlier, there are different dimensions which influence the access to

primary healthcare service. However, it is difficult to decide which dimension poses more importance

across different socio-economic groups of people. For instance, geographic distance or cost of service

might not be a problem for affluent group of people who can afford their own vehicle. Quality of

service might be more relevant for them. On contrary, cost and long travel distance might be the

prime barrier for poor people to access such service.

This type of analysis can vary greatly depending on the spatial extent or scale of analysis chosen.

Variation in result of analysis gained from different scale of data is known as Modifiable Areal Unit

Problem (MAUP). According to Schuurman, Bell et al.(2007), researches on inequality context have

given significant attention toward the construction of socio-economic inequality indicators; however

less attention is paid in addressing the influence of scale. Stafford, Duke-Williams et al. (2008) stated

that the estimation of health inequalities in specific areas are determined by the way in which the area

boundary is defined. Effect of scale, in observing variation can be analyzed by comparing results of

similar analysis conducted for the same study area, which is divided into different smaller boundaries

separately. In Stafford, Duke-Williams et al. (2008) a metropolitan city is divided in three

neighbourhood boundaries separately for analysis. The area boundaries were selected based on

predefined census wards; physical features like rivers and main roads; and according to

socioeconomic homogeneity of residents. Although the result shows small difference (around 3%) in

this case, due to certain limitations like limited sample, authors clearly state the relevance of scale in

related studies.

Therefore, the research problem to be addressed in this study is to develop an appropriate method to

measure variations in access to primary healthcare considering different dimensions in access.


3

Further, how to ensure if the level of variation in primary healthcare obtained from aggregated data

comprehends the result obtained from disaggregated one. Comparing the results obtained from large

scale aggregated data with disaggregated data shows the consistency and reliability of results from

different data sources.

1.3. Research Objectives

The main objective of this study is to evaluate access to primary healthcare, to observe if variation

exists in any dimension of access and if it meets the need of population.

The main objective is divided into three sub-objectives:

a. To develop an appropriate method to measure different dimensions of access to primary

healthcare to evaluate overall access.

b. To study how the planning for primary healthcare is done in context of Yogyakarta and to see

if existing situation matches with the planning standard.

c. To illustrate if information on socio-demographic and access to healthcare obtained from

aggregated data (census) correspond to the results obtained from disaggregated data (primary

household data).

1.4. Research Questions

For each sub-objective the questions are proposed to answer:

Questions for sub-objective 1

i. What are the appropriate methods to quantify and measure different dimensions of access?

ii. Do all dimensions have equal importance across different socioeconomic group of people?

iii. What is an appropriate areal scale for evaluating variation in access to PHC?


i. How is the planning for healthcare done based on health policies and national standards?

ii. Is the existing situation of access to healthcare in accordance to policy standards?


i. How different scale of analysis affects the result in mapping socioeconomic and service

variation in PHC?

ii. Does the result of analysis based on aggregated data, for larger spatial boundaries, matches

the actual situation within those areas?

1.5. Research Framework

The conceptual framework in Figure 1-1 shows the factors in macro level and dimensions in access,

considered to measure the existing situation of access to primary healthcare in DIY. The contextual

planning system and standards in health policies should be understood well prior evaluation of service

provision. Demographic information is relevant in this evaluation, as the need and perception on

primary healthcare service varies depending on the physical and socioeconomic condition of area as

well as population. Therefore various dimensions of access to primary healthcare are considered in

this study. By exploring these issues, dominant problematic factors causing variation in access to

4

PHC in context of study area can be found. This will highlight the important dimension in access to

be improved, in order to achieve better access to PHC to all.

1.6. Structure of Report

Chapter 1 - Introduction

This chapter gives an overall idea about the research structure starting with a brief background of

study. It comprises the research problem, main objective followed by several sub-objectives, research

questions to each sub-objectives and the conceptual framework of the study.

Chapter 2 – Literature Review

This chapter presents a brief review on definitions and concepts of access used in other empirical

studies which helped conceptualize access and its measures for this study. Detail definition of

different dimensions and conceptual framework in evaluating access to PHC adapted in this study is

explained. Variation in dimensions of access is reviewed in relation to the broader issue of equity.

Different methodologies used to quantify and measure access along with methodological problems has

been discussed.

Figure 1-1: Conceptual Framework

Problem Identification

- Dominant factors affecting access to primary healthcare

- Variation in different dimensions of access across different socioeconomic groups

- Planning and policy context

- Consistency in aggregated and disaggregated data

Suggestions and Recommendations

- Focusing dominant problematic factors under dimensions of access to further

improve existing situation of access to primary healthcare

Planning Context

Healthcare

planning

system

Planning

standards for

primary

healthcare

service

Demographic Data

Population

density

Heterogeneous

socio-economic

society

Access to Primary

Healthcare

Availability

Accessibility

Affordability

Acceptability

Adequacy


5

Chapter 3 – Description of Study Area

This chapter provides a general description of the physical, demographic and socioeconomic

condition in the Province of Yogyakarta (DIY). An overview on the effect of urbanization and

decentralization in Indonesia over provision of public services is discussed briefly. Healthcare

planning system, approaches in development of healthcare service and national policy standards for

access to healthcare in context of DIY is explained.

Chapter 4 – Methodology

This chapter describes the methodology to be carried out in this study. Detail information on research

design, data requirement, source of data and information and framework of analysis is provided. This

chapter is divided mainly into four parts; pre-fieldwork, fieldwork, post fieldwork and data analyses.

Preparation of questionnaires for household survey, base map, study area selection and sampling

strategy is explained in pre-fieldwork phase. Methods applied in carrying out ground survey,

interviews and data entry is reported in fieldwork phase. Data preparation, processing and analysis

formed the main part of post fieldwork data analyses. Various analytical methods in obtaining

answers to the research questions are described in this chapter.

Chapter 5 – Results

This chapter contains the results of analysis of variation in dimensions of access to primary

healthcare. This chapter highlighted the dominant factor or dimension affecting the overall perception

of individual on access to healthcare services. Findings of the comparative analysis between existing

situation and health policies in DIY are presented along with the results obtained consistency analyses

between aggregated and disaggregated data.

Chapter 6 – Discussion

The results of analyses, carried out to address all research questions, are discussed comprehensively in

this chapter. Critical discussion over strength and limitations of this study is also presented.

Chapter 7 – Conclusions and Recommendations

This chapter provides an overview on the concept, approaches and methods used in order to obtain

answers to all research questions of this study. Chapter concludes by highlighting main findings of

this research and further recommendations for future improvement of this study.

6

2. Review on Access to Primary Healthcare

This chapter comprises of review on definitions and concepts of access to different healthcares. It

gives detail explanation about different dimensions and conceptual framework in evaluat ing

access to primary healthcare services applied in this research. It also provides an overview on the

healthcare utilization as well as the quality of care, commonly used in evaluating health outcomes.

Issue of equity in access to healthcare has been discussed which help in clarifying the concept of

service variation and inequity in this research. This chapter further explains methodologies used

in measuring the dimensions of access by developing subjective and objective indicators. A brief

review on GIS based measures of access and methodological problems has been discussed. Issues

of different scale effects in analyses are discussed followed by conclusion of the entire chapter.

2.1 Introduction

Primary healthcare has been a major topic for many health studies, research and discussions of

medical care since decades. Increasing attention has been focused on this concern by medical

educators as well as state legislators in order to improve the supply of primary practitioners. However,

definition of the term „primary care‟ varies. Obtaining consensus on priorities or an agreement on the

content itself, has been a cumbersome task (Parker, Walsh et al. 1976). As stated in Bagheri,

Benwell et al. (2005) primary healthcare is an important step in providing „health for all‟, and is

widely acknowledged as a universal solution for improving population well-being by World Health

Organization and UNICEF. Guagliardo (2004) states that primary care is an essential form of

healthcare for maintaining population health as it is relatively inexpensive and easily delivered.

Author further explains that it is most effective in preventing disease progression on a large scale if

they are adequately provided in space.

Access to primary healthcare is considered as one of the indexes in achieving the goal of „health for

all‟ and it has different definitions depending upon different contexts. Different definition and

concept of access to primary healthcare is explained in the next section.

2.2.1 Definitions and Concepts of Access to Primary Healthcare

Access to primary healthcare can be seen from a broad perspective making it difficult to give a

precise definition. Efforts to conceptualize and measure access have varied depending on different

circumstances and context of study. Being a concern of social welfare, studies on access to healthcare

service have been carried out by different professionals like geographers, public health sector,

anthropologist etc. Thus the definition varies accordingly depending on the study approach. “The

most basic problem in defining „access‟ is that it is both a noun referring to potential for healthcare

use, and a verb referring to the act of using or receiving healthcare” (Guagliardo 2004, p.2). This

might create confusions due to overlapping of understanding between physical presence of primary

healthcare facility and ability and willingness of people in obtaining the care.

A more specific definition refers to ability to “secure a specified set of healthcare services with

certain level of quality, subjected to a specified maximum level of personal inconvenience and cost,

while in possession of a specified amount of information” (Oliver and Mossialos 2004, p.656). The

term „specified‟ in this definition, makes it easier for policy maker to define access depending on

specific circumstances for different places depending on the availability of resources to finance

healthcare. It can be summarized that general definition of access can guide the planner and policy

maker towards important factors, such as relevant range and quality of service, inconvenience,


7

disutility, cost, information dissemination etc, to be considered. If adopted and implemented properly,

this can serve as a standard against which existing access can be compared. Therefore this helps

policy maker to observe how improvement can be made or to check if they are improving access over

different area and population.

As various factors and issues should to be considered while defining access to primary healthcare,

some common factors can be grouped into different dimensions. Study of various dimensions and

factors affecting access to primary healthcare can further clarify the definition of access.

2.2.2 Dimensions of Access to Healthcare Service

Many authors have mentioned about different factors that impede access to primary healthcare. Due

to the complexity of access concept, it is important to look at each factor separately, even though they

are interrelated. Factors like availability of medical personnel, convenience to achieve health

services, actual use rates, service use in relation to some standards of need and consumer satisfaction

level with services has been highlighted by many studies, while exploring the overall access to

primary healthcare service. Dimensions of access differ with different geographical, socioeconomic

and cultural settings. Healthcare insurance, service cost, physical distance to reach the service, lack

of transportation, capacity of facility to serve the need of patients, indirect cost apart from health

insurance like travel cost, socio-cultural factors like race, language, gender etc and service quality

issues are some issues in access to healthcare service as documented in Millman (1993). Author has

grouped these issues into three dimensions; structural, financial and personal. Similarly, Penchansky

and Thomas (1981) has grouped those issues, termed as barriers, into five dimensions: 1. availability,

2. accessibility, 3. affordability, 4. accommodation and 5. acceptability; ‘5 A’. The first two

dimensions are spatial in nature. Availability refers to the total number of service from which user can

make their choice. Accessibility is related to travel impedance (time or distance) between spatial

location of user and services. The last three dimensions are non-spatial, related to cost, service quality

and cultural factors. Obrist, Iteba et al.(2007) also adopted above mentioned five dimensions in

clarifying the concept of access to healthcare but the term „accommodation‟ is replaced by

„adequacy‟ while explaining if patient‟s expectation towards quality of service and personal treatment

is met by the facility. The structural dimension in Millman (1993), possess the factors mentioned in

availability and accessibility, financial dimension covers the affordability and personal dimension

includes issues of adequacy and acceptability.

The concept of „5A‟ is used in this study while evaluating existing situation of access to primary

healthcare. Each dimension is explained and divided further into simpler quantifiable and measurable

form.

1. Availability refers to the extent to which a system provides facilities (which is the structural

form) and services (which refers to the process) that meets the needs of people (Campbell, Roland et

al. 2000). More than simple doctor-patient ratio, availability further deals with access to specific

gender of medical personnel, for example female general practitioner or nurse; access to medical

stores, laboratory or other equipments etc. Campbell, Roland et al. (2000) stated that organizational

access can be seen as sub-component of availability. This means even if people have adequate

physical access to the facility, there might be other factors creating barriers like length of time in

getting appointments, waiting time in before getting treatment or sometimes language barrier with the

facility professionals.

2. Accessibility is most commonly related with the geographic location of patient to the

location of facilities. Measures like spatial distance, travel time, mode of transportation used to reach

the facility, type to road network etc are considered assessing physical accessibility of people. Large

8

number of studies like (Talen and Anselin 1998; Black, Ebener et al. 2004; Bagheri, Benwell et al.

2005; Amer 2007; McGrail and Humphreys 2008) has been done on physical accessibility to public

facilities and mostly healthcare service. “Accessibility relates to the ability of people to overcome the

friction of distance to avail themselves of services at fixed points in space” (Amer 2007, p. 31). Study

of physical accessibility mostly incorporates three components: people; activities or service and mode

of transport to link them. The framework in Moseley (1979) states that accessibility varies according

to the characteristics of each of these components and it is influenced by the relationship between the

socioeconomic character of people, users, and spatial dimensions. More elaborative concept on

accessibility incorporates one more component, the temporal component along with the three

mentioned above. This deals with the moment of time at which the service is available or at which

people are able to participate. For example, the opening hour of the primary healthcare facility and

working hour of people in this study.

3. Affordability by its name itself refers to the financial component. There might be adequate

number of health facilities or medical personnel in an area, spatially close to the needy population.

But, these facilities might not be affordable to them. In such case, people tend to go to other facility

than the one closer to them, which provide subsidized rate or cheaper service. This dimension looks

into direct cost like doctor‟s fee as well as indirect costs like travel and medical costs that have affect

on overall access to healthcare. Other factors like possession and coverage of health insurance, public

supports such as subsidized rate provided for certain group of people like low income or elderly

population, are also incorporated in this dimension.

4. Acceptability deals with the cultural also religious factors of people. Factors like age,

gender, education level, race or ethnicity determines the level of acceptability of service provision to

large extend. For example, if the service available is socially acceptable by people like gender issue,

if people have some religious or cultural preferences towards choosing certain healthcare facility or if

the service provider and people use a common language to communicate. This also depends upon the

personal perception of people that might vary within a same religion or gender. Beliefs and

expectations of different groups of people should be considered while evaluating this dimension.

5. Adequacy is seen from two ways in this study: quality of service provided and personal

treatment by the service providers. Opinion about the medical treatment whether people trust the

medical ability provided by the facility or not, if they are satisfied with the quality of service or

personal behaviour of all facility personnel right from the point of entry to facility for example person

at reception till the end of medical treatment by doctors and laboratory personnel.

Acceptability and adequacy being subjective matter, it becomes complex to define a specific standard

or threshold line. These can be looked from the people‟s feeling, preference and perception on related

issues.

A list of empirical studies related to access and its dimensions, along with methodological approach,

that are reviewed for this study are summarized in Table 2-1.

These five dimensions can also be related to the concept of potential and realized access to healthcare

used in Andersen, McCutcheon et al. (1983) and Bagheri, Benwell et al. (2005). The potential access

looks at the ability of needy people in gaining healthcare in presence of healthcare facility in space. It

is important to understand the difference between “having access” and “gaining access” to healthcare

while evaluating the access to the service. Gulliford, Figueroa-Munoz et al. (2002) clarifies the

distinction between them, as the former may be the result of the availability of services and the latter

one refers to ability of individual to overcome financial, organisational and socio-cultural barriers to

utilise healthcare service.


9

Some researchers tend to relate potential access with characteristics of the population (such as their

household income, education level, their attitudes toward medical care etc) and the characteristic of

the delivery system (like the type of organization, distribution of medical personnel and types of

facilities). Such characters aid in the understanding of meaning of access toward healthcare for

different socio-economic group of people. The realized access deals with actual utilization of

healthcare services. To simplify, potential access consider the factors, influencing the ability to enter

facilities before receiving the service and realized access deals with the perception of people towards

different aspect, such as quality of service, cost or personal treatment, after receiving the service.

In this study, availability, accessibility and affordability along with other socioeconomic

characteristics of people are looked from potential access to primary healthcare. Realized access

includes issues related to acceptability and adequacy. Once factors and dimensions of access are

determined, a general framework can be formulated to observe a stepwise sequence in evaluation of

access to primary healthcare.

Author and

year

Study Context and

Area

Dimensions in Access

Mea

sure

s

Operationalization /

Methodology

Av

ail

ab

ilit

y

Acc

essi

bil

ity

Aff

ord

ab

ilit

y

Acc

epta

bil

ity

Ad

equ

acy

/

Acc

om

mo

dati

on

Penchansky

and Thomas

(1981)

Concept of access in

health policy

United States

√

√

√

√

√

S / O

- Factor Analysis

- Multiple regression

- Correlation coefficient

Andersen et

al. (1983)

Access to medical

care

United States

√

√

√

√

-

S / O

- Descriptive statistics

- Correlation coefficient


- Factor Analysis

Fosu, G.B.

(1989)

Access to healthcare

in urban areas of

developing societies

Ghana

√

-

√

√

-

S / O

- Descriptive statistics

- Pearson correlation


- Standardized regression

coefficient

Guagliardo

(2004)

Spatial accessibility

studies in urban

areas

United States

√

√

-

-

-

O

-Two-step floating catchment

area

- Gravity model

- Kernel density method

(Black,

Ebener et al.

(2004)

Physical

accessibility to

health care

Honduras, Central

America

√

√

-

-

-

O

- Correlation analysis

- Regression analysis

Bagheri et

al. (2005)

Spatial accessibility

to primary

healthcare

New Zealand

√

√

-

-

-

O

- Drive time and least cost path

analysis model (network

analysis) in Arc Info 9.1

10

Author

and year

Study Context

and Area

Dimensions in Access

Mea

sure

s

Operationalization /

Methodology

Avail

ab

ilit

y

Acc

essi

bil

ity

Aff

ord

ab

ilit

y

Acc

epta

bil

ity

Ad

equ

acy

/

Acc

om

mod

ati

on

Omer (2006) Spatial equity

regarding physical

accessibility to

urban services (park)

Israel

√

√

-

-

-

O

- Proximity

- Container measure

- Minimum distance

- Moving average index

- Correlations income, religion

and obtained service area

Obrist, Iteba

et al. (2007)

Exploring and

improving access to

healthcare in

resource-poor

countries

√

√

√

√

√

S / O

- Outcome in terms of

health status

- Patient satisfaction

and equity survey

- Multivariate Analysis

Amer

(2007)

Spatial equity in

urban health services

planning

Tanzania

√

√

-

-

-

S / O

- Statistical analysis

- Pearson correlation coefficient

- Two step Cluster Analysis

- ANOVA

- GIS-based „what if‟

- Flowmap

Yamauchi,

Chowdhury

et al. (2007)

Spatial coordination

in public good

(education and

health facilities)

allocation Indonesia

√

√

-

-

-

S / O

- Descriptive statistics;

cumulative percentage,

distribution

- Non parametric regression

McGrail and

Humphreys

(2008)

Measuring spatial

accessibility to

primary care in rural

areas

Austrailia

√

√

-

-

-

O

- 2SFCA

- Network Analysis in ArcView

9.1, closest facility

Table 2-1 : Empirical references to dimensions of access

Summarized overview of the empirical literatures on evaluation of access and its dimensions to

public services (majority on healthcare service)

Note:

= Included in the literature

= Not included in the literature

S / O = both subjective and objective

O = Objective

Looking at this table it can be said that limited number of studies has considered non-spatial

dimensions like affordability, acceptability and adequacy while conceptualization and

operationalization of access to primary healthcare. Measuring spatial components like physical

accessibility and availability have been a common approach in evaluating access to healthcares.

- √


11

2.3 Conceptual Framework for Access to Healthcare

Different authors have developed their own models and framework in assessing access to healthcare

depending on the way they define it. Most of the literatures in healthcare give emphasis on the

outcome of service in the form of health status and equity on service. However access to healthcare

has been an important part in their framework. In the model of Millman (1993), dimensions of access

to healthcare is the first step which is followed by service utilization, quality and efficiency of service

and finally by the health outcomes. The framework adopted in explaining health access by Obrist,

Iteba et al.(2007) is shown in Figure 2-1. This framework shows the combined effect of macro level,

i.e. the policies and different healthcare services, and micro level, i.e. livelihood assets of people and

need of healthcare, on the dimensions of access. Obrist, Iteba et al.(2007) explained about livelihood

assets comprising of five popular capitals: human capital (like education level, skills or local

knowledge); social capital (social networks or affiliations); natural capital (natural resources like

land, water or livestock); physical capital (physical infrastructure like road and facility complex,

mode of transportation, facility equipments etc) and financial capital (income level, subsidies or

health insurance). Further it is explained that the availability of these assets is influenced by less

controllable factors like economy, political state, technological advancement or natural factors like

climate or hazards like flooding, draught or epidemics. These factors are referred as the vulnerability

state. Once access is gained, then the framework follows the similar pattern like Millman (1993).

This framework comprises both supply (health services) and demand side (health seeking behaviour)

and it places access in the broader context of livelihood assets. Interaction between; healthcare

services and organizational policy or institutions responsible to govern the services and; livelihood

assets that people can use during vulnerability contexts determines the extent to which access is

reached along the five dimensions.

With reference to these literatures, a conceptual framework to evaluate five dimensions of access for

this study was prepared, which is shown in Figure 2-2. Like in Obrist, Iteba et al.(2007), effects of

macro and micro level are observed on the dimensions of access to healthcare services. Factors in

Figure 2-1: The health access framework

[Source: Obrist, Iteba et al.(2007)]

Note: Red highlights the main focus of this literature

12

each dimension should be measured to evaluate access. Once access is gained and service utilization

is ensured, then people‟s perception on quality of care and satisfaction over each dimension can be

used to analyse most influencing dimensions over access and their relationship with the overall

satisfaction level of people.

Figure 2-2: Conceptual framework to evaluate access to PHC

Although this study is limited to evaluation of dimensions of access, an overview on the service

utilization and quality of care can be relevant to develop a clear concept on total healthcare system.

This will allow to compare the perception of individuals on quality and satisfaction over different

dimensions with those documented empirically in previous studies.

2.3.1 Healthcare Utilization and Quality of Care

Once need for primary healthcare is realized and access to service is acheived, then the quality of

service provided can be checked by evaluating the healthcare utilization rate. This is carried out in

Conceptual Framework to Evaluate Access to Primary Healthcare

A

C

C

E

S

S

Accessibility

Availability

Affordability Acceptability

Adequacy

5 A

Interrelation

Hierarchy of planning organization,

Health planning agencies, policies

and processes

Healthcare services

Hospitals, government healthcare

services, private clinics, traditional

healers, medical stores and others

M

a

c

r

o

L

e

v

e

l

Physical and Socio-economic

condition of people

(Livelihood Assets)

Need for Primary Healthcare M

i

c

r

o

L

e

v

e

l

Ensure access to primary healthcare

Findings

Patients‟ satisfaction with factors in

each dimension

Perceived variation in access

Feeling of equity in treatment

State of access at micro level in relation

to policies at macro level

Results

Influencing factors of dimensions in

access

Relation between satisfaction level with

each dimension and overall satisfaction

with access to PHC


13

the „Finding’ section in Figure 2-2. Obrist, Iteba et al. (2007) mentioned that, improved access and

healthcare utilization should be combined with good quality of care in order to get positive health

outcomes. The outcomes can be observed by measuring health status, patients‟ perception towards the

level of satisfaction and equity. Quality in healthcare is multidimensional. Quality can be evaluated

in respect with some predetermined standards (normative) or can be judged by more subjective way

like measuring patient‟s satisfaction level. For example, Maxwell (1992) have point out that results

obtained from the people opinion on satisfaction can differ from the result obtained by evaluating

technical efficiency of facility.

Campbell, Roland et al. (2000) defined quality of care from two principal dimensions; access and

effectiveness. Access, as already mentioned, confirms if people are getting adequate primary

healthcare at time of need and effectiveness confirms how effective is the care obtained. Quality of

care is defined as “the ability to access effective care on an efficient and equitable basis for the

optimization of health benefit/well-being for the whole population” (Campbell, Roland et al. 2000,

p.1617).

Quality of care has been discussed in more elaborative form in Maxwell (1992). Author has point out

some important components while defining the quality of care.

Effectiveness: ensures if the treatment provided is the best available in a technical sense and

evaluates the result of treatment.

Efficiency: compares if the healthcare output is maximised for a given input or vice versa.

For example comparing unit cost among different health facilities providing similar services.

Relevance: observes if the overall pattern of services is the best possible considering the

needs and expectations of population.

Equity: looks into issues like relative fairness in treatment. It ensures if some people or group

of people are being dealt less favourably.

These issues prove to be important while evaluating existing state of primary healthcare in my study.

Individuals‟ perception on issues like service quality, trust towards the ability of service providers and

feeling of equity has been addressed while evaluating their satisfaction level with existing primary

healthcare provision.

2.3.2 Equity in Access to Primary Heathcare

Achieving equity in access to primary healthcare has been a central objective to many health care

systems. Most governments have declared that citizens should get universal and equal access to good

quality of primary healthcare. The issue of equity should be addressed at macro level in the

framework shown in Figure 2-2, while developing policies for health care planning.

“The concept of equity refers to the degree to which services or amenities are distributed in an equal

way over different areas as well as economic, ethnic and political groups, with appropriate

consideration given to the needs of special groups such as children and the elderly” (Omer 2006,

p.254). Depending on context, policy makers give priority to different groups of people to secure

equal access to primary healthcare. For example such groups are commonly defined by income; social

status; geographical location; education level; ethnicity; gender; lifestyle etc. According to Andersen,

McCutcheon et al.(1983) equity of access is obtained when services are provided on the basis of

people‟s need. This refers to proportional equity based on needs. Authors state that inequity exists if

factors like race, income level or insurance coverage etc are important predictors of access.

14

Equity can also be seen from social aspect. According to Martinez (2005), to reveal the fact that

inequalities within cities really do matter, it is essential to consider an approach from social justice

perspective. Smith (1994) considers that justice involves treating people fairly, which in distributive

justice means that whatever is being distributed should go to people in the right quantities. Social

justice is concerned with the question of who gets what where and how, and more precisely who

should get what where and how.

Equity in this study is related with the variation in primary healthcare service received by people

across different socioeconomic classes. Further equity can be observed from peoples‟ perception on

quality of service and personal treatment they receive from healthcare facilities. Comparing similar

issues between different service providers, for instance government and private clinics, can further

clarify the equity concept in this study.

Once concept is clear and dimensions of access are determined, an appropriate method should be used

to measure each dimension in order to evaluate the state and (un) equal access to primary healthcare.

2.4 Measuring Dimensions of Access to Primary Healthcare

The need to measure each dimensions of access individually has been recognized in literatures in

order to evaluate the overall access to primary healthcare either with respect to certain norms or from

user‟s perspective. As the main objective of this study is to evaluate access to PHC from five

dimensions adopted from literatures like Penchansky and Thomas (1981) and Obrist, Iteba et

al.(2007), developing an appropriate method to measure each dimension becomes the main

methodological problem in this research. To measure each dimension of access, they can further be

divided into simpler quantifiable components. Developing quantifiable indicators of each component

is a commonly adopted method in measuring access to healthcare. To clarify the concept of indicator,

a brief definition, advantages and types of different indicators are explained in the following sections.

2.4.1 Developing Indicators to Quantify and Measure Dimensions of Access

Andersen, McCutcheon et al.(1983) have stated that there have been a number of summaries of the

research on the indicators that correlates with evaluation of healthcare service which should be

considered in various approaches while measuring access. Developing such indicators allows to

measure variation, if exists, in any dimension of access to primary healthcare across different socio-

economic groups. And if variation is found to be comparably high, then the indicators further enables

to measure inequalities in different aspect of access. As mentioned in Martinez (2005), advantage of

using indicators to measure inequalities is that they can communicate in a simple way while detecting

and quantifying such variations. Indicators can also be used in monitoring the area of priority for

policy intervention. The main functions of indicators can be summarized as to simplify complex

phenomenon, quantify them to measure and to communicate the results. Depending on functions and

purpose of policy, indicators are classified into three groups in Parnell and Poyser (2001) which is

also explained in Martinez (2005).

• Descriptive or baseline indicators: describes the existing situation of some system or process. They

are the initial data collected for each variable.

• Normative or target indicators: Once area of need is identified, target indicators help to set the

expected target or goal to be achieved. They allow evaluating and comparing the existing condition

with certain standard normally defined by policies.

• Performance or outcome indicators: enables to check if the targeted goals have been achieved from

policy side and also allows observing users satisfaction level with the obtained result.


15

This study however is concentrated in developing descriptive indicators that exhibits the present

condition of access to PHC across different socioeconomic groups of people, considering different

dimensions of access. Development of indicators is focused on the needs, experiences and opinions of

people (users) towards the existing access to PHC. This helps in highlighting the problematic issues or

area of need to draw policy attention. Based on the concept of five dimensions of access, indicators

were developed referring to the questions raised towards each dimension in Obrist, Iteba et al. (2007).

To address the research question for sub objective 1, indicators were developed to distinguish

different socio economic groups of people to see variation in dimensions of access. Various household

characteristics can be summarized as indicators in order to come up with different socioeconomic

classes. List of socioeconomic and access indicators developed for this study are presented in

Appendix A with references from number of empirical studies in related field. These indicators can

be continuous data like income, crowding, employment dependency or age dependency and

categorical like education level, housing condition or ownership of assets. Detail description and

rationale behind all indicators developed for this study purpose are presented in Appendix B.

The evaluation of access can be seen from two ways: subjective and objective perspective. Therefore

subjective and objective indicators can be developed separately to measure them. The following

section will explain more about these indicators.

2.4.2 Subjective and Objective Indicators

Being a matter of social concern, policy makers in healthcare planning have been interested in both

objective and subjective approach in policy making process. Hence, developing subjective and

objective indicators becomes important in this approach. Large number of literatures like Veenhoven

(2002), Das (2008), Foo (2000) etc have stated the importance of both subjective and objective

indicators while evaluating the quality of life, where healthcare has been an important domain of

study. As explained in these literatures, objective approach focus on measuring „hard‟ facts like

income of family, education level, distance travelled to reach healthcare, cost paid for the service etc.

The subjective approach on the other hand measures „soft‟ matters like the people‟s opinion or

perception on some issue, their satisfaction level with income or the healthcare services. Subjective

indicators express individual‟s evaluation of objective matters. Subjective indicators are normally

measured using Likert scale. Likert scale is a commonly used scale in research surveys and

questionnaires related to subjective measures and respondents state their agreement to a level of

statement. For example, 5-point Likert scale for satisfaction can range from: very satisfied, satisfied,

neutral, dissatisfied and very dissatisfied. However there is no such defined range in Likert scale. For

instance Turksever and Atalik (2001) used a 4-point Likert scale in evaluating subjective quality of

urban life, while Das (2008), Foo (2000) and Ibrahim and Chung (2003) applied 5-point Likert scale.

Both of these approaches hold their own importance in evaluating social problems and developing

new policies. As mentioned in Veenhoven (2002), objective indicators provide information about the

actual state of social problems and the effects of attempts to solve these problems. Such information is

of indisputable nature or can be called objectively true which enables rational social action. Despite

of many controversies against subjective indicators, like being irrational that hampers scientific

management, Veenhoven (2002) stated that it is important in social policy making process for number

of reasons such as:

1. Need of social policy to consider peoples‟ psychological satisfaction and their support, as policy is

not limited to mere material matters. Therefore, subjective indicators are required to achieve these

subjective goals.

2. To evaluate the progress or outcome of policy intervention from subjective measurement.

3. Satisfaction level and preferences of people gained directly from people can better indicate

comprehensive quality of services provided.

4. The distinction between „needs‟ and „wants‟ of people for certain service can be measured with

subjective indicators.

16

In this study, objective indicators give information about the physical state of access to healthcare

like spatial distance to healthcare service, travel time, waiting time, direct and indirect cost to be

paid for service etc. Subjective indicator on other hand will provide information on the mental state

of people towards all these issues along with their satisfaction level with over all access to healthcare

service. Subjective indicators are concerned with individual‟s evaluations of various aspects of their

healthcare-seeking experience like the convenience to get service, cost, service provider behaviour,

or overall quality of the care they receive. For example, there might be a case when a person can

physically access a healthcare with ease but he might not be happy with the quality of service or

personal behaviour of facility personnel. There might be discrimination in medical as well as personal

treatment for different socio economic class of people which leads to dissatisfaction among certain

group of people. This kind of information related to personal feeling is gained from subjective

indicators.

It is important to observe link between results obtained from subjective and objective approaches, as

there are some contradictory conclusions about the relationship between these two approaches. For

example, Das (2008) exhibits a weak relation between subjective and objective approach in

measuring quality of life. However, Foo (2000) and Ibrahim and Chung (2003) recommend to use

both indicators in order to complement the limitation of specific indicators.

2.4.3 Analyzing and Measuring Indicators

The conceptualization and the measurement of both indicators have been a major concern in

healthcare research. According to Andersen, McCutcheon et al. (1983) multiple regression techniques

have been applied to analyze a range of potential and realized access indicators. Objective measures,

which normally refers to utilization rates, can be measured in different ways for example, simple

proportion of people visiting and not visiting a healthcare facility within a certain period of time or

total volume to service consumed. While analyzing subjective measures, some researchers have used

satisfaction indicators as a determinant of utilization and others relate it as a consequence of

utilization. Roghmann, Hengst et al. (1979) states that satisfaction measures are content oriented and

its validity is generally assumed or assessed through correlations with other variables like their

willingness to change the service providers if applicable. In such studies, regressions were computed

to predict satisfaction from utilization and vice versa. Other statistical methods like descriptive

statistics, multi regression, correlation and factor analysis can be applied to analyze the nature and

relation among different types of data collected on various aspects of study (Field 2005). Lahelma,

Martikainen et al. (2004) used logistic regression analysis to calculate inequality indices for health

service. Martinez (2005) stated that factor analysis and multivariate statistical techniques are the

most commonly used techniques in social research which are preferable approaches in measuring

socio-spatial variations. Twp step cluster analysis or K-mean methods enables to obtain different

clusters of socio economic classes using continuous as well as categorical variables as input data

These statistical analyses are relevant for this study for exploring and analyzing collected data on

both subjective and objective measures of access. These analyses will enable to obtain relation

between different dimensions of access and helps to validate the result obtained to evaluate overall

access to healthcare service.

Along with statistical techniques, application of Geographical Information Systems (GIS) has a great

importance in analyzing the observations and displaying results. Next section explains about the GIS

application in related type of research.


17

GIS based Measures to Access

Increasing advancement in GIS in health organisations, together with the availability of data, has

supported studies related with developing measures of access to healthcare services. Black, Ebener et

al. (2004) stated that GIS is suitable in measuring spatial accessibility to healthcare as they enable

researchers to input, store, manage, manipulate of both spatial and attribute (textual) data, analyze

and visualise spatial information. Literatures like Bagheri, Benwell et al. (2005), Guagliardo (2004)

and many more have explained about measuring spatial dimensions of access; availability and

accessibility. Measuring straight line distance (Euclidean distance) and creating thiessen polygons are

some simple methods in assessing physical access. For more detail analysis, sophisticated network

analysis can be done calculating travel time, distance and mode of transport in GIS. Bagheri, Benwell

et al. (2005) calculated driving time to primary healthcare (destination point) from patients‟ house

(origin point) to apply least cost path analysis model using network analysis in Arc Info 9.1.

Higgs (2005) reviews number of literatures highlighting the use of GIS-based measures in exploring

the relationship between geographic access, utilization, quality of healthcare service and health

outcomes. These studies explore the spatial configuration of healthcare delivery system along with

service quality measures; role of transport system used to reach the care service for different socio

economic class of people and the characteristics of people or the area where they reside, seeking

healthcare in measuring access to healthcare service.

GIS software provides spatial analysts that offer excellent tools for spatial data management and

visualization. Amer (2007) has used GIS to identify and visualize trend like socially progressive or

regressive pattern in the distribution of healthcare facilities. GIS in combination with other software

like Flowmap is useful in identifying suitable location of facilities for certain number of facilities to a

defined population or territory. This can ensure the optimal distribution of service in space

considering the concentration of demand (people). Amer (2007) mentions that more sophisticated

GIS-based analytical techniques , for instance GIS-based „what if‟ in his study, can be used to

evaluate and support improvement of healthcare service in terms of spatial equity and efficiency

which can further support strategic planning of urban health services delivery.

As this study focus on spatial and non spatial components of access equally, a simple proximity

measures can be applied to examine the influence of spatial dimensions on total access to primary

healthcare service. Further observation on how peoples‟ perception of proximity to primary healthcare

facilities can be related to actual proximity to validate the conclusion of this research.

2.5 Methodological Problems to be Resolved in Evaluating Access

Main methodological problems mentioned in the academic literature related to equal access are

twofold. According to Oliver and Mossialos (2004) the first problem is to get consensus in the

development of specific definition of healthcare which will enable healthcare policy makers to make

policy that is more consistent in providing suitably (dis)proportionate access across different groups of

people with different level of needs. This refers to providing appropriate unequal access for unequal

need. The second problem is to develop an appropriate method to measure access. This requires the

formation of standard policy that specify explicitly issues like: 1) the minimum relevant range which

can be referred as a benchmark while evaluating the quality of healthcare services, 2) considerable

level of convenience for all healthcare users, 3) margin of cost to be paid in obtaining basic

healthcare and 4) the minimum amounts of information that people should have to get the advantage

of healthcare services.

However, getting total consensus in defining access to healthcare is a cumbersome job, as is it not

feasible to consider and fulfil the need of total population. Different individual might have different

18

perception and priorities about the concept and dimensions of access. More over developing universal

standard to evaluate access can be impractical as it is highly determined by the geographical location,

political condition to some extend and socioeconomic as well as cultural characteristics of users.

Homogeneity of area, defined by above mentioned factors should be considered while developing

social policies. Therefore, developing different standards for different homogeneous areas,

considering little tradeoffs to minority, might be appropriate for measuring and evaluating

dimensions of access in that particular area.

Another important methodological issue in related research in access is the effect of geographical

scale of analysis in the results. The affect of scale of analysis over results is discussed in the next

section.

2.5.1 Effect of Scale in Analysis

Spatially aggregated data often provided by statistics office are commonly used by researchers to

investigate the contextual determinants of health. Most studies use census areas as the geographical

units for convenience as detailed population data are available in this unit. Spatial extent or scale at

which the analysis is done has an effect on their values and results. According to Stafford, Duke-

Williams et al. (2008) results of analysis using aggregated data vary according to the selection of

areal unit which is well-known as Modifiable Areal Unit Problem (MAUP). Schuurman, Bell et

al.(2007) states that the complexity of scale in research analysis is of critical importance as large

number of deprivation indices on health outcomes are being developed for policy implementation.

Researchers use different indicators at different spatial scales to find area of service deprivation. This

allows to observe the relationship between socio-economic difference and variation in healthcare

service. This enables appropriate classification of high-risk populations, or the areas of deprivation to

inform policy makers. “MAUP refers to the problem that occurs when inferences, based on spatial

analysis, change when the same data are analyzed using either variations in administrative zoning or

through different scales” (Schuurman, Bell et al. 2007, p.596). The two main components of MAUP

mentioned in literatures are the scale effect and geographic boundary constraint.

Number of studies (Carstairs 1981; Cockings and Martin 2005; Haynes, Daras et al. 2007; Stafford,

Duke-Williams et al. 2008) has stated that automated zone design techniques can be an option to

counter the affect of MAUP. This helps to control the design of zoning system in order to create

robust aggregation of spatial information for the intended analysis to be undertaken. In automated

zone design, different zones are created by automated means in which zone boundaries are controlled

by statistical design rules. Zone design technique can be used to create zones by maximizing the

internal homogeneity of variable like physical or socio economic condition within each zone. A

potential application could be the testing of hypotheses of causal links between variables. For

example, if it is hypothesised that health condition in socially deprived area is worse than in well off

areas, one could aim to create zones which are homogeneous in terms of social deprivation. If the

hypothesis is true then strong resultant correlation could be expected between the independent

variable, deprivation in this case, and dependent variable, health condition, for the newly created

zones. Such zones might be created considering approximately equal population size at different

scales (Cockings and Martin 2005) or internal homogeneity in terms of environment or social

characters (Carstairs 1981; Haynes, Daras et al. 2007). Such zones are able to demonstrate stronger

relationship between variables than census units, and that larger areas produced stronger relationships.

Similar techniques were used in the 2001 England and Wales census to produce homogeneous output

areas (Martin, Nolan et al. 2001).

Haynes, Daras et al. (2007) grouped 1991 Census enumeration districts (EDs) using similar material

deprivation values , like housing type, in the zone design. Homogeneity of variables in each zone can


19

be measured by the intra-unit correlation (ICC) which can be calculated by dividing the variance

component between zones by the total variance. Stepwise multiple regressions can be used to identify

significant predictor variables.

From the review of these literatures, inappropriateness of spatially aggregated census data and ward

boundary for specific analysis like health condition or service can be assumed. It can be concluded

that automatically designed aggregation might reflect the actual situation for accurately than the pre-

existing administrative area boundaries. Rogerson (2006) explains about the inconsistency in results

of statistical analyses obtained from aggregated and disaggregated data. It is noted that “correlation

coefficients tend to increase with the level of geographic aggregation when census data are analyzed.

A smaller number of large geographic units tend to give a larger correlation coefficient than does an

analysis with a larger number of small geographic units” (Rogerson 2006, p. 165).

Effect of scale is relevant in this study, as it aims to check consistency in results obtained from census

and primary data. This will ensure the reliability of aggregated data over disaggregated one for

particular studies like socioeconomic variability and access to PHC.

2.6 Conclusion

This chapter reviewed the general definitions and concepts used to describe access to primary

healthcare service by various authors in different geographic context. As there is no such universally

accepted standard in defining and measuring access, commonly used dimensions of access was

explored which further developed into a model in measuring and evaluating overall access to

healthcare provision. This contributed in development of the conceptual framework for this study.

Although this study concentrates on the evaluation of different dimensions of access to healthcare,

review on different phases and aspects of healthcare outcome such as livelihood focus, healthcare

utilization, quality of healthcare and health related equity issues was very useful to develop an overall

idea on the primary healthcare system.

Access to healthcare can be measured by two approaches: subjective and objective approach. These

can be measured by developing subjective and objective indicators for each dimensions of access

respectively. In this research, indicators are the significant tools which simplify the dimensions in

such a form that they can be quantified and measured.

Overview on different statistical and GIS applications in related researches are relevant for

developing methodology for this study. Relation and dependency of variable in access can be

obtained from different statistical analysis. GIS can be effectively used: to explore spatial

characteristics of study area, in estimating proximity to healthcare facilities, to explore the spatial

variation in access to healthcare and finally to visual interpretation of the results.

The review showed the importance of access to healthcare, regardless to geographic or socioeconomic

context, in social policy making process. The output of this study is expected to highlight factors

influencing the access to primary healthcare and to display how spatial as well as social variations

exist in this regard.


21

3. Study Area Description and Healthcare

Policies

This chapter presents a brief description on the study area and the planning process for

healthcare. The description includes the geographic location, surface area, administrative units,

and influence of decentralization, socio-demographic condition and other characteristics of the

Province of Yogyakarta. Overview on the procedure and indicators used by government planning

agency to develop policy standards related to health care service is also explained.

3.1 General Description of Study Area

The province of Yogyakarta is located in the South of Central Java Island. It is the second smallest

province out of 33 provinces, after Jakarta in Indonesia with the total area of 3185 km2. It is

surrounded by the province of Central Java and bounded by the Indian Ocean on south. A distinct

character of this province is that it still has its pre-colonial monarchy embedded in the administrative

structure. Yogyakarta is the only province that is led by kingdom Kraton Yogyakarta, besides the

formal government. Due to this reason it is known as the special region of Yogyakarta, Daerah

Istimewa Yogyakarta or DIY in Indonesian. Apart from monarchy there is democratically elected

legislative body, Regional People's Representatives Assembly known as Dewan Perwakilan Rakyat

Daerah. Javanese is the main ethnicity with 97% of total population. Majority of people in DIY are

city are Islamic with 91.8% followed by Christian (7.9%), Hindu (0.2%) and remaining Buddhist

(0.1%).

3.1.1 Demographic Condition

Based on the result of National Socio-Economic Survey of Indonesia in 2005, total population in the

province was recorded to be 3,281,800 in which 51% was female and 49% of male. Sleman regency

has the largest population and Kulon Progo is the least populous. The population density in 2005 was

1030 person per km2 with the growth rate of 1.88 percent. Between 1990 and 2000, Sleman regency

had a growth rate of 1.45% accounting for the highest rate in the entire province. This phenomenon is

supported by high population density in the municipality and compared to the others during that time.

Figure 3-1 presents the population growth trend in the province from year 1971 to 2000.

The highest population density at present is in Yogyakarta city that is around 13,000 people per km2

with area around 1% of total DIY area. In contrary, the density of Gunung Kidul regency is the lowest

with around 470 people per km2 with area coverage of 46% of the total DIY area. Figure 3-2 displays

the population density per village in DIY according to figures in census 2005 and Table 3-1shows the

population density of DIY per regency. The province has experienced significant growth in terms of

population in over 3 decades. Based on the result of National Socio-Economic Survey, the highest

percentage of DIY population by age is productive (16 – 60 years old) people. About 14% of the total

population are old, above 60 years. This shows that population of DIY tends to have a higher life

expectancy.

22

General exploration of demographic attributes of the Province of Yogyakarta is done in Arc GIS.

Census data obtained from the Statistics office of Indonesia (BPS) for the year 2005 was used as it is

the latest date for data availability. Information from census data is available on village level for the

whole province. Some general demographic information for five regencies in DIY are listed in Table

3-2.

Legend

Regency

Village

Population

Density

(person per

sq.km)

Figure 3-1: Population growth trend in

DIY (1971 – 2000)

Source: Statistics Indonesia, 2008

Figure 3-2: Population density map of DIY

Data source: Census, 2005

Regency Bantul Gunung Kidul Kulon Progo Sleman Yogyakarta

State Rural Rural Rural Rural Urban

Total population 800569 747782 447695 895408 515976

% of population in DIY 23.5 22 13 26 15.5

% of area coverage in DIY 16.5 46 18 18.5 1

Male (%) 48.5 48 49 49.5 51

Female (%) 51.5 52 51 50.5 49

No. of household 214558 178936 110867 236776 102716

No of districts 17 18 12 17 13

No. of villages 75 144 88 86 45

Table 3-2: General demographic characteristics of DIY Source: Statistics Indonesia, 2005

Regency Area (km2) Population Density

2003 2004 2005 2006

Kulon Progo 585 640 640 660 635

Bantul 505 1600 1610 1,625 NA

Gunung Kidul 1,485 460 460 470 460

Sleman 575 1635 1640 1660 1755

Yogyakarta 32 12,025 12,445 12,935 13,600

Total 3,185 1000 1010 1030 787.15a

Table 3-1: Population density per regency in DIY from years 2003 –

2006 Source: Statistics Indonesia Note: NA = Not available a = Not including Bantul regency


23

3.1.2 Landuse and Economic Activities

DIY has been divided into two major land types: rural and urban. All four regencies have been stated

as rural. Only city of Yogyakarta is defined as urban city by Indonesian Department of Public works.

Population density and economic activities are the main criteria in defining it as urban. Apart from

rural and urban land use, some areas are stated as preserved areas that include natural resources like

mountain range and water catchment areas. Figure 3-3 shows the main land uses in DIY.

The economic structure in DIY is boosted by public service sectors like educational institutions and

tourism. Agriculture, trade, and industry also have some contribution in economic development.

Tourism sector has the major contribution to the GDP (Gross Domestic Product); however,

agricultural sector employs the most people. Communication and transport sectors are gaining

prominence over years which aid to development to large extend. Main economic activities in urban

land use are trade and service while less populous rural areas are predominant by agriculture.

However steady conversion of agricultural land to build up form can be observed in the periphery of

Yogyakarta city, especially in Sleman on South. Shift from agriculture to other economic sectors has

lead to tremendous growth in such sectors. This growth is attributed to huge amount of migrant

population from areas outside the province. High rate of rural urban migration has resulted in high

population density with widespread unemployment and lower living standard. Unemployment is

problematic but with the rate fluctuating between 5–8%, which is comparatively lower than in other

regions in Indonesia. These phenomenon leads to a heterogeneous society, dominantly in the city and

its surrounding areas.

3.1.3 Administrative Units

There is a descending level of administrative subunits in the government administration process.

There are twenty seven provincial level units which are divided into districts (kabupaten), sub-

districts kecamatan) and further into villages (desa or kelurahan). Village is the lowest tier of the

administrative hierarchy. The nation is centrally governed from Jakarta since independence, from

where national level decisions like line of authority, budget allocation, personnel appointments are

done. Regional and local governments enjoy little autonomy. Their role is mostly administrative like

implementing policies and regulations. They essentially serve as subordinate administrative units in

local level, which support the functional activities of Jakarta based departments.

The DIY has four districts also known as regencies; Sleman, Bantul, Gunung Kidul and Kulon Progo

and one city, the city of Yogyakarta which is the capital of the province. City of Yogyakarta is

located centrally, surrounded by Bantul and Sleman on the South. Gunung Kidul is the largest

regency with surface area of 506.80 km2 where as the City of Yogyakarta has the smallest surface

area of 32.5 km2. The administrative units of DIY are shown in Figure 3-4. Rural regency has a

regent and mayor in case of city of Yogyakarta, who holds the authority for the local development of

public service sectors like education and healthcare facility. Next administrative unit is the sub

district which also has a head person who has certain level of power in development and decision

making processes. The sub district heads are directly accountable to the mayor or regents. Similarly

there are village heads and village head office in each village, where village development plans or

other service related problems are discussed before proposing to the higher administrative levels.

24

3.2 Health Policies and Planning Systems

Overview on the decentralization in Indonesia is useful for this study as the planning process and

policies for healthcare service follows the regulations after decentralization. Significant influence of

decentralization can be seen on public service sector including healthcare.

3.2.1 Decentralization

Decentralization is the devolution of political decision making to local governments and

communities. The Indonesian Ministry of foreign affairs has defined decentralisation as “a means to

hand over political, financial and administrative authority from central to local (regency/city)

governments, so that the government can facilitate and guarantee better public services for the

people.” Indonesia experienced decentralisation in 1999 with the passage of certain laws on Regional

government (Law number 22/1999) and on fiscal balance between the centre and the regions

(Number 25/1999). This restructured the administrative framework by providing local government

the responsibility for planning and provision of public services. Central government is however

responsible to monitor and evaluation such development processes. Decentralization is viewed as a

positive development towards stronger association and transparency between people and public

service developments. One advantage is that authority at local level is more aware with the ground

situation and has greater knowledge about the service related problems. So they can meet the needs of

people more efficiently, also being easily accountable by local people. By the distribution of power,

better representation of local people is assumed to be achieved ensuring maximum public

participation in development processes. It is further believed to enhance effective distribution of

public services. Empirical studies on public services in Indonesia, documented in Yamauchi,

Chowdhury and Dewina (2007), have demonstrated that physical accessibility to public services like

school and hospitals has improved during decentralization period. But, per capita availability of

school and local medical healthcare, puskesmas, have decreased over time. The authors mentioned

that despite of the coordination in spatial allocation of such services, availability is coming up as a

problem due to mobility of endogenous people and higher growth rate. These phenomena partially

cancel the advantages of the coordinated efforts on public service allocation. There are some

drawbacks of decentralization in public service delivery. Yamauchi, Chowdhury and Dewina (2007)

states that investments in public service can be biased depending on local income and endowment

which complicates the coordination of investment decisions across different socioeconomic

communities. However it falls outside the scope of this research.

Figure 3-4: Administrative

boundary Data source: Census, 2005

Legend

Figure 3-3: Land use of DIY Data source: Census, 2005

Legen

d

Built forms

Regency

Forest

Agriculture

Green land

Water body


25

3.2.2 Health Policies and Strategies

The government of Indonesia has been redeveloping policies to improve the health service throughout

the nation. One of the policy visions in the line is the creation of “Healthy Indonesia 2010”. As

health is considered as a shared responsibility, the policy forces the Ministry of Health and Social

welfare to develop collaborative relationships with other parties involving all strata of community,

other related government agencies and private sector. As stated in WHO (2009) the “Healthy

Indonesia 2010” is developed with the main goals:

1. To initiate and lead a health orientation of the national development

2. To maintain and enhance individual, family and public health along with improving the

environment

3. To maintain and enhance quality, accessible and affordable health services

4. To promote public self-reliance in achieving government health

This study is related more to third and fourth goal. In the mean time, with the introduction of two new

Acts on Local Governance (Act No. 22/1999) and on Financial Balance (Act. No 25/1999), the

implementation of decentralization policy in Indonesia was emphasized from 2001. This empowered

provinces and regencies with autonomy in formulating policies and decision making in health service

considering local needs. It is believed that the service can be provided more effectively and urgent

problems of the area can be resolved faster. Collaboration between the National Health Development

and decentralization policy agreed on four paramount strategies serving as the pillars in formulating a

Strategy for National Health Development, which are:

a) Concept development of health

b) Professionalism

c) Community health maintenance assurance (Jamkesmas) and

d) Decentralization

The third strategy is relevant for the scope of this study as it deals with the „affordability‟ dimension

in evaluating access to primary healthcare. It focuses on community health insurance programs like

Jamkesmas. The aim was to guarantee equity in access to health services and the service quality. This

program was based on the concept of adequacy, equity, efficiency and effectiveness.

Indonesian government has been trying to increase the healthcare for the poor people. According to

the document provided by the health agency in Yogyakarta, in year of 2009 Department of Public

Health raises the number of poor people who receive subsidy for health insurance premium from 36

million people (2005) to 60 million people (2009). Health insurance premium was also increased

from 2.1 trillion (Indonesian Rupiah) to 3.7 trillion. By this increment, 60 million poor people are

expected to receive free healthcare by bringing the insurance card to government hospital and local

government clinic, puskesmas. As this policy is explicitly formulated for poor people, it is important

to understand how „poor‟ is identified. For social stratification, numbers of socioeconomic indicators

are developed by the Planning Office at regency / municipality level in context of local area and

living standards. The indicators includes attributes like income, employment status, physical

condition of house, nutrition, possession of consumer goods like telephone, television, motor bike, car

etc. According to the planning officer in Yogyakarta, the criteria for deciding different

socioeconomic class of people does not focus on mere income, as considering the possible affect of

26

inflation in future. It is assumed that with certain level of income, one might not be able to maintain a

similar kind of living standard in future, the way they are living at present, if the market price rises.

These indicators are developed in reference to the census data. The central Bureau of Statistic (BPS)

is responsible to provide data needed for similar planning procedure and for implementation. BPS

data is used also for the financial allocations to districts and in estimating „quotas‟ for poor people.

The social data produced by BPS is collected through surveys like National Social economic survey

(SUSENAS). Such survey is designed in order to collect social population data which is relatively in

wider scope. Collected data includes information on education, health/ nutrition, housing /

environment, criminality, social culture actions, consumption, income of households, possession of

vehicle type and other consumer goods and household welfare level. A team is set at sub-district level

for the survey in identifying poor households. When a household is considered to be poor after

evaluating with those indicators, then it will receive a letter known as „Surat Keterangan Tidak

Mampu‟. With this letter of poor the household receives free or highly subsidized rate at public

services. With the possession of this letter, one becomes eligible to get a health card known as „Katu

Sehat‟. The local authorities provide the lists of qualifying individuals and districts are allocated

quotas for health cards, based on the estimated percentage of poor people residing in that district.

Such cards cover the service cost of both outpatient primary care in puskesmas and free treatment at

hospitals (generally at third class public hospitals).

A nominal rate of Rp. 5,000 (about US dollar 0.55) per month per card holder is set as the standard

contribution rate. At present there are different types of free health insurance premium like Jamkesda

(district level health insurance for poor), Jamkesmas (civil health insurance for the poor), Askeskin

(Indonesia health insurance for the poor) etc depending upon their area and employment status. The

main function of these insurance is the same as previous health card that is to provide subsidized

health service and free in case of primary healthcare.

The national standards for healthcare service related to this study are presented in Table 3-3.

Service Area Indicators Service Standards Quality

Scope Services

Health care

facilities

-Distribution of

healthcare facilities /

coverage of health

services

For neighbourhood

unit population

should be less than

30,000

-At least 1 unit of

puskesmas / sub

district

-1 unit of health post

(posyandu) / village

-1 unit medical post

/ 3000 people

-1 unit of child and

maternal healthcare

(maternity hospital /

BKIA) / 10,000 to

30,000 people

-1 unit puskesmas /

120,000 people

-1 unit hospital /

240,000 people

Location of

facility should be

in centre of the

sub-districts /

districts

Should be located

in clean area far

from disease

sources , garbage

dumping sites and

pollutions

Table 3-3: National standards for health service, Indonesia

Source: Standard guidelines of minimum service level, 2001

The health policies seem to give priority to the spatial distribution and service cost. The report

(Moeloek 1999) from the Ministry of Health documented that in spite of adequate number and

distribution of heath facilities, the health services are still below standards in terms of service quality

as it is far from peoples‟ expectation. Another problem mentioned is regarding the uneven distribution

of health service personnel, inadequate educational quality and unbalanced health manpower


27

composition which results in longer waiting time by patients in healthcare centres and low service

performance. Although, there is not any officially documented policy regarding this issue,

information related to waiting time in healthcare facilities was obtained from the interview with head

of health planning agency and puskesmas personnel during fieldwork. According to them, the

indented optimal waiting time is considered to be around 30 minutes. However, these issues need

further policy attention.

3.2.3 Health Organization System

After implementation of decentralization, 349 regencies and 91 municipalities (also referred as

district/regency/city) became the key administrative units. One level down in administrative

hierarchy is the sub district, which has at least one public health centre, puskesmas headed by a

medical doctor. It is usually supported by two or three sub-puskesmas, generally headed by nurses.

The medical doctors occasionally visits the sub-puskesmas for instance once in a month for

monitoring. Most of the puskesmas are equipped with four wheel ambulances for emergency cases

also to support mobile health service to provide service to underserved populations in urban as well as

remote rural areas. At village, the smallest administrative unit, there should be at least one integrated

family health post. These health posts are established by community people with the help of medical

assistant from health centres. The aim is to provide preventive healthcare service also to promote

healthcare awareness. Midwives, both professional and traditional, are deployed to the villages

commonly for maternal and child healthcare.

The hierarchy of organization of health system starts with the Ministry of Health at the national level.

It is then followed by health planning agencies known as dinas kesehatan at province and further at

district levels. Then puskesmas are established at sub district level which is followed by sub

puskesmas, midwives and health posts at the lowest administrative unit (village). The flowchart of

organizational structure of health system is shown in Figure 3-5.

MOH level from central to peripheral level

Ministry of Health, Central level

Departemen Kesehatan (DEPKES)

Provincial health office, Province level

Dinas Kesehatan Propinsi (Dinkes)

District level health office, District level

Dinas Kesehatan Tingkat Kabupaten / Kota

Sub district level health center

Pusat Kesehatan Masyarakat (Puskesmas) tingkat Kecamatan

Village level health center

Posyandu

Sub Health Center Village Midwife Clinics Integrated Health Post

Figure 3-5: Organizational Structure of Health System in

Indonesia

28

3.2.4 Planning Process for Healthcare Service

Decentralization has provided a platform to support more community participation and involvement

of different stakeholders in the development and planning process for public services. Both local

communities and government organizations are often involved in the process of decision making for

public goods investments and service allocation. This has encouraged the bottom up planning system.

When people realize a need of certain healthcare service, villagers coordinated by the village office

with a village head approach the health office at sub district level and then to district level with a

proposal of investment project for that service. Then the district health office set a team of survey for

the evaluation of validity of the need of that project. In presence of many project proposals, the

district office selects a good project based on priority analyzed from the survey and thorough

analysis. Then it allocates fund to the selected project, received from central government allocated

for that region. However this process has some problems like difficulty in coordination and lengthy

decision making period. As both local communities and government are involved in the decision

making process for different projects, gaining consensus is a very difficult and lengthy process. And

even with the government‟s effort to coordinate investments across communities in their jurisdiction,

it cannot be guaranteed that the investments will be well coordinated to gain equal access of the

service by all people. Hence attention is given in the coordination for an effective and faster project

approval, planning and implementation to meet the demand of people.

Conclusion

This chapter provided general information on demographic condition in DIY such as population

density, population growth trend, urban rural land use types and hierarchy of administrative

boundaries. Such information was relevant for getting familiarize with the study area and also

important in selection of areas for primary data collection. Brief explanation on health policies and

planning process helped in understanding the healthcare system at macro level. This was important as

one of the objectives in this study was to compare existing situation of access at micro level with the

policy standards in the area context.


29

4. Research Methodology

This chapter contains the methodological approach to address the research questions of this study.

Answers to the research questions are obtained by following a method that has five stages. In the

first stage, literature and empirical studies of access to healthcare provided a firm base to develop

initial concept of methodology. Research design is developed to determine the required data,

source of data and methods of obtaining them. The second stage is pre-field work preparation for

the fieldwork. General study about the study area along with the overview of available secondary

data and information was carried out. In relation with the research question and possible

availability of secondary data, questionnaire was prepared for household survey and interview

with primary healthcare facilities. The field work of 28 days was carried out during which all

relevant primary and secondary data was collected in the third stage. All collected data is

processed and checked for its consistency in the fourth stage which is the post field work stage.

The final stage is the data analysis. Methods for analysing both subjective and objective attributes

are applied to study the existing state and variation in access to primary healthcare in the study

area.

4.1 Research Design

The research design of this study is shown in Table 4-1. It presents details on data required, source of

data collection, methods to be applied to obtain answers for research questions to achieve each sub-

objective. The research design also includes the type of analysis carried out in this study

4.2 Fieldwork Preparation

A base map of study area was prepared with the available spatial data on administrative boundary and

geo-referenced Google image. Various indicators were developed to quantify and measure five

dimensions of access identified after literature review. Indicators developed under each dimension of

access are presented in Appendix B along with description and rationale. List of all relevant

information and household questionnaire for primary data collection was prepared accordingly. The

questionnaire consisted of two major sections: first was the general information related to household

and their socio-economic characteristics, second part focused on factors related to the access to

healthcare facilities. This section included type of healthcare facilities respondent visited and their

perception on satisfaction with various attributes of access to the facility. The questionnaire

comprised questions on both subjective and objective attributes of access. From objective attributes

answers for travel time, travel distance, cost, income level etc were obtained. Subjective attributes

included personal perception of level of satisfaction with various dimension of access to health care

facilities.

The questionnaire was developed in a simple structure which can be understood easily by

respondents. Questions related to subjective attributes and ranking options were translated in the local

language, Bahasa in order to be filled by the respondent themselves. An interview format was

prepared to obtain information from health facilities personnel.

30

Sub-

objectives

Research

Questions

Required data Data source Methods Analysis

1 - What are the

appropriate

methods to

quantify and

measure different

dimensions of

access?

-Do all

dimensions have

equal importance

across different

socioeconomic

group of people?

- What is an

appropriate areal

scale for

evaluating

variation in access

to PHC?

Literatures,

Empirical studies

Census data

Socio-

demographic

characteristics

Spatial factors

(geographic

location and

distances)

Individuals‟

perception

BPS –

Statistics of

DIY

Household

survey

Center for

Transportation

and Logistic

Studies,

Gadjah Mada

University

(Pustral,

UGM)

1. Developing

indicators for

socioeconomic

characteristics of

respondent

2. Developing

indicators for each

dimension of

access

3. Measuring

perceived level of

satisfaction with

each dimension

and overall access

4. Comparing

perception on

satisfaction level

between different

economic classes

of respondent.

5. Synthesising

indicators to

developing

summary scores

for dimensions

Descriptive

statistics; mean,

range, standard

deviation

percentage

count

Two step cluster

analysis

GIS; Proximity

measure

Chi square test

Correlation

coefficient

Coefficient

matrix

Plotting

summary scores

in multi-

dimensional

charts

Standardized

residuals

3

2

-How is the

planning for

healthcare done

based on health

policies and

national

standards?

-Is the existing

situation of access

to healthcare in

accordance to

policy standards?

Health policy

and planning

standards for

healthcare

facility in DIY

Demography data

Number of

existing

healthcare

facilities

Spatial factors

(geographic

Planning for

Health

Department

(Dinas

Kesehatan)

Regional

Development

Agency DIY

Province

(Kepatihan

Yogyakarta)

BPS –

Statistics of

Studying planning

process for

healthcare service

in DIY

Comparing

predefined

standards with

existing situation.

Ratio of

population to

healthcare

facilities and

Total number of

doctors

Distance

proximity

measure

Evaluating

individuals‟

perception over

policy measures


31

location and

distances)

DIY and their

implementation

4

3

-How different

scale of analysis

affects the result

in mapping

socioeconomic

and service

variation in

PHC?

-Does the result

of analysis based

on aggregated

data, for larger

spatial

boundaries,

matches the

actual situation

within those

areas?

Census data:

Aggregated

socio-

demography

data;

information on

access to

healthcare;

spatial data

Disaggregated

data on similar

issues from

primary source

BPS –

Statistics of

DIY

Pustral, UGM

Household

survey

Comparing

aggregated census

data with

disaggregated

household data.

Evaluating

consistency in

statistical analysis

Descriptive

statistics;

percentage

count

Coefficient of

variation

Correlation

coefficient

Table 4-1: Research Design

4.2.1 Study Area Selection

For the selection of study area, units of administrative boundaries were considered. The province of

Yogyakarta consisted of five regencies (districts), 77 sub districts and 438 villages. For this study,

village was an appropriate unit of observation since it was the basic unit of consensus-building for

bottom up approach in decision makings. The scale of analysis was chosen also in concern to the

secondary data to be obtained, as data were available on such administrative boundary. Certain

criteria were developed for selecting sample villages for detail study. Population density,

socioeconomic heterogeneity and spatial location of villages were considered for selection.

Percentage of prosperous household figure presented in census and visual interpretation using Google

image was used to determine heterogeneity in socio economic condition. All regencies in DIY

followed similar health policies formulated at the province level, regardless to different socio-

demographic structure in rural and urban regency. For instance, according to one of the health

policies, there should be at least one puskesmas in each sub district. However it did not mentioned

about the geographic area or population size of such sub districts. As area and population of sub

districts in rural and urban regency has vast difference, a comparative analysis of access to PHC

between rural and urban area was intended in this study. City of Yogyakarta being the only urban

regency, it was selected purposively. Two villages Tegalpanggung and Kricak were selected from

city of Yogyakarta. Tegalpanggung had the highest population density in the whole province and

located in the centre of city of Yogyakarta, hence was selected. Kricak was located at extreme

northern boundary of the city with comparatively higher population density than other villages

located in boundaries. Reasons to consider geographic location was to observe variation in access to

PHC, if any, in highly crowded city core and growing city fringe. Another spatial characteristic of

these villages was the river and major road network defining the physical boundaries. Dense informal

settlement was found along the river bank in west boundary of both villages. In contrast to this, more

formal basically commercial activities were found along main road lines. Such characteristics were

assumed to provide the possibility of exploring heterogeneous group within each village. Villages

32

were selected also considering the geographic area coverage and percentage of population covered by

those villages in relation to the respective sub districts they belonged to. Selected villages had the

largest area and population size among other villages within that sub district, hence was considered to

be representative for that sub district. This criterion was checked as the third sub objective of this

study was related to analysis of scale effect. Hence for comparative analysis between aggregated data

for sub districts and disaggregated data at village level, most representative village were chosen.

Sleman was selected as rural regency as it had highest population density among four rural regencies

in DIY. Village Tridadi was chosen from Sleman considering higher population density and

heterogeneous mixture of formal and informal settlement observed from Google image.

Administrative support provided to carry out survey and interview in related health organizations in

this village also played some role in the selection. After selection of villages, sampling strategy was

designed to carry out household survey. The main objective in this strategy was to collect data which

was representative of the population in village as far as possible.

4.2.2 Sampling Strategy

Depending on time constraint and limited financial resource, total sample size for this study was

determined to be around 300 households from three selected villages. This was about 3 percent of

total number of households residing in all three villages. Therefore proportional number of household

from each village was calculated considering the same percentage. Random sampling was applied for

the selection of households within those villages.

After the purposive selection of villages and determination of sample size, random sampling was

applied for the selection of household for survey. Digitization in ArcGIS was done to define the

cluster of built forms within each village. The digitized pockets excluded the major road and vacant

lands as far as possible. The built form in two villages in city of Yogyakarta was more dense and

clustered together. Therefore, grid of 100 meters by 100 meters was laid over the digitized pockets

and then calculated numbers of random points were created over grid boxes in ArcGIS. These points

were pre determined for each village, i.e. 3 percent of total household in the village. Digitization and

random points were also laid for the third village. However, grid was not applied due to the sparsely

located smaller pockets of built forms. Number of sample points within each digitized pocket was

decided according to the area covered by those pockets. Location and random points of sample units

for household questionnaire is shown in Figure 4-1.

4.3 Field Work

The required data for the study was collected from two main data sources; primary and secondary

sources during fieldwork.

4.3.1 Primary Data

Household survey was carried out in the selected villages along with three surveyors in their local

language. Objective data and individual perception about different aspect of the primary healthcare

facilities was collected through structured questionnaire from 28th September 2009 to 9th October

2009. Prior to the real household survey, the questionnaire was well explained to the surveyors and

required translation of certain portion in questionnaire was done. A pilot survey was carried out in

one study village to confirm the surveyors understanding of the questionnaire, also to check the time

length required to complete one questionnaire. The criterion for household survey was that the

respondent must be the head of family or the spouse of the head of family. This criterion was set so

that the respondent should be able to give the detail information about the whole household members,


33

for example overall income level. It was assumed that the head of family or spouse can reflect the

perception or feeling of all other household members.

100m X 100m grid Tegalpanggung

Kricak

Tridadi

Figure 4-1: Location of surveyed household in

villages Note: Image obtained from Google Earth-Pro, acquired in 2007

34

Ten questionnaires were filled in pilot survey. The time required to complete one questionnaire was

about 20 to 30 minutes. Certain changes were done after the pilot survey in order to further clarify the

questions. Surveyors were given training in reading map over laid with satellite image, in order to

find the spotted house on image for survey. In case of unavailability or unwillingness of respondent in

the spotted house or if the spot is on land use type other than residential, the survey was carried out in

the nearest house from the spot. Second part of the primary data collection was interview with

personnel at two health planning agency, Dinas Kesehatan at City of Yogyakarta and Sleman regency

and with medical staffs at 5 public healthcares, Puskesmas, in three sub districts containing the

sample villages. For this purpose, an official request letter was issued from the University of Gadjah

Mada and a permission letter from the municipality was obtained. Then mentioned health planning

agencies and puskesmas were visited along with a local translator in the appointed date. Purpose

behind the interviews at health planning agencies was to gain understanding about the health planning

process and related policy structure in study area. And purpose of interviews at puskesmas was to

understand type of health services being provided and to query about common problematic factors, if

any, in access to healthcare reported by patients. Some photographs from fieldwork are displayed in

Figure 4-2 .

Figure 4-2: Data collection and fieldwork observations

(a) Household survey with the help of local surveyors

(b) Socioeconomic heterogeneity with in village (Kricak)

Well of households Poor informal settlement along river bank

(c) Data collection from health planning agency and public healthcare facility, puskesmas

Health planning agency Waiting area in Puskesmas


35

4.3.2 Secondary Data

Census data for the whole province of Yogyakarta for year 2005 was obtained from the Center for

Transportation and Logistics Studies, Gadjah Mada University. This data was provided by the

Statistics office of the province. Spatial data like administrative boundary of study area for the whole

province along with road network data was also obtained from the same organization. Census

contained number of socio-demographic information at village level for the whole province.

Knowledge on the planning system for health facilities was gained from the visit to regency planning

office and health agency. Few supporting documents on numbers of different healthcare facilities and

medical doctors were also acquired. Table 4-2 displays the list of secondary data collected during

field visit.

4.4 Post Fieldwork

Data collected from primary source, questionnaire and interviews, is processed, checked for

consistency and entered in digital form. After entering, every 10th household data was cross checked

with the filled questionnaire, to ensure the accuracy in entering. As most of the collected documents

were in local language, Bahasa, required translation was done. As the census data contains a huge

number of attributes related to socio demographic information, infrastructures, agriculture and other

public services, only limited attributes relevant to this study were selected and translated in English

for further use.

4.5 Challenges During Fieldwork

Some difficulties and constraints were faced before and during the data collection.

Lack of sufficient data and information on socioeconomic condition and health facilities

hindered the selection of villages and finalization of questionnaire before field visit.

Due to absence of high resolution satellite image, preparation of base map for sampling for

household survey took time. Processing high resolution image from Google pro and geo-

referencing and preparing mosaic from small pieces of images was time consuming.

Spatial location of healthcare facilities was not obtained as per expectation. Location of five

puskesmas, in three sub districts containing selected villages, and hospitals being visited by

respondents were spotted in Google image with the help of secondary document collected

from health planning agencies and with the help of local spatial knowledge. Personal visit

was made to the spotted health facilities for conformation.

Language barrier was another difficulty during fieldwork. As a consequence, survey had to be

done with the help of local surveyors and translators.

Documents obtained from planning and health agency was in their local language.

Translation required extra time and effort.

Sample size had to be limited due to time and financial resource constraint.

Due to the Muslim festival „Idul Fitri‟ fieldwork had to be stopped for a week. During the

festival period all related organizations were closed for interviews. Also due to unavailability

of local surveyors even household survey could not be carried out.

36

Type of Data Description Data condition Source

Spatial data Administrative boundary;

Province, regency, sub-district

and village

GIS data (vector) Pustral, UGM

Road network GIS data (vector) Pustral, UGM

River GIS data (vector) Pustral, UGM

Demography data Population data, geographic

socioeconomic data

Excel

Document (hard copy)

Pustral, UGM

BPS- Statistics of DIY

Health Facilities

Number of health facilities per

village (public hospitals,

puskesmas)

Excel


Regional development

agency DIY

(Kepatihan Yogyakarta)

Number of medical doctors per

village

Excel


Health Agency

(Dinas Kesehatan)

Puskesmas

Standards in health policy and

information on regional health

insurance

Document (hard copy) Health Agency

(Dinas Kesehatan)

Table 4-2: Collected data from secondary sources

4.6 Data Analysis

4.6.1 Socioeconomic Stratification of Sample Households

Analysis of socioeconomic strata of respondent was relevant, as the major part of this study was based

on the subjective perception of respondent on dimensions of access to PHC, which was assumed to be

influenced by their socioeconomic characteristics. Also socioeconomic attributes such as possession

of letter of poor and government health card, explained in section 3.2.2, were important as being

directly related to the dimension affordability. Number of socioeconomic attributes collected from

household survey was combined together in order to categorize each sample households into different

socioeconomic groups. Description and rationale for each socioeconomic indicators used is presented

in (Appendix B).

In this study different attributes of household characteristics were given equal consideration in

evaluating their socioeconomic status rather than limiting to a single variable like income. Amer

(2007) explained that using single variable like income or education to distinguish social strata

oversimplifies the result and cannot reflect the actual reality. The author also highlights the negative

possibility of using scoring method in which variables are assigned certain weights based on some

assumption, which are later combined to come up with one value as the final result. This approach has

some drawbacks, as the process of assigning weights is preconceived and subjective approach.

Therefore to avoid such possibilities, Amer (2007) used a multivariate and exploratory stratification

approach, Two-Step Cluster analysis. This analysis is adapted in this study for similar purpose.

Two-Step Cluster analysis is an exploratory tool designed to reveal natural groupings within a data

set. The algorithm used in this process offers number of beneficial features like 1. ability to create

clusters using both categorical and continuous variables, which is suitable for the data set in this

study; 2. it enables automatic selection of the number of clusters based on similarities and

dissimilarities within and between clusters. It also allows user to specify the number of clusters if


37

needed and 3. it can efficiently analyze large data set. Clusters can be characterized based on the

descriptive statistics provided for each input variable. After classification, GIS is applied to visualize

the spatial distribution of households from different socioeconomic classes within each sample

villages, to see socioeconomic heterogeneity at village level.

4.6.2 Measuring Dimensions of Access to PHC

Descriptive statistics was used to measure both objective and subjective indicators developed under

dimensions of access to PHC. The first three dimensions of access to PHC; availability, accessibility

and affordability, was measured from objective as well as subjective approach by developing

indicators. The remaining two dimensions; acceptability and adequacy, being subjective in nature

only subjective indicators were developed to measure perception of respondent on related issues.

During household survey, questions were asked about objective and/or subjective aspect on each

dimension of access to PHC. The subjective perception over each indicator and overall satisfaction

level with access to PHC was measured using 5 point Likert scale that ranged from very satisfied to

very unsatisfied. Statistics like mean, standard deviation and minimum – maximum range was used to

compare variation in objective indicators such as travel distance, time and waiting time in health

facilities between different villages, health facility users and socioeconomic classes. For travel

distance, Euclidean distance was measured in GIS between sample household points and visited

public health facilities as reported by respondents. Descriptive statistics, i.e. cumulative percentage of

respondents is calculated to compare the subjective satisfaction level with each indicator and over all

access to PHC.

For categorical data, cross tabulation table was used as it is a basic technique for examining the

relationship between nominal or ordinal variables. It also offers a test of independence and measures

of association between such variables using Cramer‟s V statistics. If the frequency of subjective

opinion in Likert was less than 5 in cross tabulation, then 5 point Likert scale was aggregated into 3.

This was done by considering „satisfied‟ and „very satisfied‟ as one value indicating positive opinion

and „dissatisfied‟ and „very dissatisfied‟ was replaced with one value indicating negative opinion.

Neutral opinion was left unchanged.

Correlation matrix was computed between levels of satisfaction with various indicators and overall

access to see the significance and strength of correlations. Factors which had highly significant strong

correlation can be considered as better predictors of overall satisfaction level with access to PHC.

4.6.3 Synthesising Indicators

Summary scores were developed for all dimensions of access by synthesis of underlying indicators, as

it is hard to deliver a meaningful message from indicators alone when looked from a broader

perspective of policy making. Developing a summary score by synthesising indicators by considering

their relative importance into a single value has been a common practice which enables easier

interpretation for policy maker. “It is the analysis of indicators against the wider context and policy

objectives that provides the added value of converting information into intelligence” (Wong 2006,

p.81). This will provide a more explicit evaluation of the state of access to PHC following the

concept developed for access in this study.

Considering equal importance of all dimensions in access, equal weights were applied to each

indicator while developing a summary score. Also as there was no reference of subject matter expert‟s

opinion on prioritization of different indicators, equal weights were applied. The concept followed

while developing indicators based on their relevance referring to literature Obrist, Iteba et al. (2007)

can fairly justify the use of equal weights for this study.

38

Box plot was used to see the nature and range of summary score distribution. Following Wong (2006)

average value of summary scores for each village, health facilities and socioeconomic classes were

plotted in multi dimensional radar chart for visual interpretation. As the result of equal weighting,

information on dominant problematic factors might get lost while developing summary scores. Hence

referring to Amer (2007), standardized residual was used to see variation in occurrence of satisfaction

level with each indicators under dimensions. Residuals give the difference between observed

frequency of cases with satisfaction and the expected frequency, specifying the deviation from the

expected average number. It helps to observe how proportion of respondents from different cases

(villages or using different health facilities) differed based on their satisfaction with each indicator.

4.6.4 Access to PHC in Relation to Existing Health Policies

One of the objectives of this study was to compare the existing situation of healthcare service in terms

of access, with health policies and planning standards in study area. This enabled to compare the

existing situation of access at micro level from people‟s perception with planning norms at macro

level. As discussed earlier in section 3.2, after decentralization, provinces and regencies were

empowered with autonomy in formulating policies for public facilities including health. Hence health

policies and standards formulated at provincial level were referred for descriptive analysis.

Information on spatial distance to health facilities and number of doctors working for public health

facilities, obtained from census data was used to prepare choropleth maps to display spatial variation

in geographic distances from villages to nearest public health facilities in DIY. Also maps were

prepared to display relative variation in population doctor ratio referring to the Indonesia‟s average

population doctor ratio obtained from Dash (2000).

Descriptive statistics, i.e. cumulative percentage, was used to compare the existing state of factors

related to access to healthcare, which are also mentioned in health policies of DIY.

4.6.5 Scale Effect in Analyzing Socioeconomic and Access Variation

To achieve answers to the research questions related to the effect of scale in analysis, a comparative

study was done between primary and secondary data source. The purpose was to explore the

consistency of the results obtained from aggregated and disaggregated data. Referring to different

techniques to minimize scale effect (refer section 2.5.1) from empirical studies (Carstairs 1981;

Hyera 2003; Stafford, Duke-Williams et al. 2008), this study was concentrated on socioeconomic

homogeneity of three selected villages to observe variation in socioeconomic characteristics and

access to healthcare within each village. Common variables related to socioeconomic and access to

PHC was selected between primary data and census for comparison.

Descriptive statistics like percentage count, mean, minimum-maximum range of the common

variables was used to see how well census figures at village level represented the actual condition

within villages obtained from primary data collection.

Variation observed at village level was compared with the variation at sub district level using similar

variables from census data. Referring to Turksever and Atalik (2001), variability in socioeconomic

attributes and access to PHC was explored in terms of coefficient of variation (CV) at sub district and

village level. Coefficient of variation was computed by dividing the standard deviation of each

attributes by its mean and it indicated absolute variation within sample villages. This will further aid

in analysing variation obtained from different sources of data. In all applicable cases, CV was

computed using aggregated village data from census to observe variation at sub districts level and

disaggregated household data for variation within each sample village.


39

Referring to empirical studies on scale effect, it was assumed that correlation coefficients tend to

increase with higher level of geographic aggregation. “A smaller number of large geographic units

tend to give a larger correlation coefficient than does an analysis with a larger number of small

geographic units” (Rogerson 2006, p.165). Hence to ensure consistency of aggregated and

disaggregated data in statistical analysis, correlation coefficient was checked between similar

variables of access from different source of data.

40

5. Perceived Access to PHC at Micro and

Macro Level

This chapter presents the results of analysis in four main sections. First section includes the

general characteristics of the sampled households, emphasizing socioeconomic attributes.

Measurements and analysis of indicators on each dimension of access to PHC is presented in the

second section. The results of the analysis of perceived satisfaction level with each dimension

including variability in their multiple indices are presented in this section. Third section evaluates

the existing state of access comparing with health policies and standards in DIY. The last section

contains the comparative analysis of common variables between census and household data to

show the consistency of using aggregated data in this type of study.

5.1 Household Characteristics and Socioeconomic Stratification

Household characteristic

General characteristics of 273 sampled households were explored to get an idea of the collected data

prior to running other analyses. Majority of respondents were female and about 75% of households

were composed of 3 to 6 members. As the survey was carried out in day time between 9 am to 6 pm

and most of the households have male as working members. As the province of Yogyakarta is

renowned for educational institutes, literacy rate was found higher than the national literacy rate

which was 71% in 2008 according to Statistics of Indonesia (BPS), also larger than the ratio for the

province itself (65% of literacy). Person above 15 years old who have attained formal education and

are able to read and write is defined as being literate by BPS. About 79 % of sample households were

found to attain education from senior high school up to university level education. The household

monthly income was classified into 5 ranges starting from 700,000 Indonesian Rupiah (IDR).

Absolute income poverty line for province of Yogyakarta was defined as 194,830 IDR per capita for

year 2008 by BPS, which is equivalent to around 21 US dollar per month. This was less than 1 dollar

per day per capita if referred to Millennium Development Goal poverty indicator. Using such

standard poverty line could not reflect the actual state of reality, as it differs from country to country.

Hence, poverty thresholds or indicators should be chosen as per the context of study area. Range of

income was designed assuming average household size of four members considering the income

poverty line by BPS. About 38% of households stated to be in the range below 700,000 IDR out of

which 65% of households have more than 4 members. The percentage of households below income

poverty line in the province was 18% in 2008 as stated by BPS. Around 67% of households reported

to have the letter of poor from the local government. Majority of the interviewed houses (52%) were

semi permanent with half non brick or wall and non plastered floor. This shows a significant

difference with the figure stated by BPS for 2008, which was 13% of semi permanent houses in the

province. Definition of each category of construction type was referred from the census questionnaire

from the BPS (presented in Appendix B). Regarding basic infrastructures like water source and

sanitation, majority of households with 58% used well water and around 77% had private toilets with

septic tanks. These figures are similar with the percentage presented by BPS for the whole province

of Yogyakarta.

From these figures, it can be said that the study area have larger proportion of poor households as

compared with the statistics for the province provided by BPS. However, it is in better state in terms

of literacy rate when compared with the provincial as well as national literacy rate.


41

Socioeconomic stratification

A number of socioeconomic indicators were used as input variables in the Two Step Cluster Analysis

to assign each household a socioeconomic class. Prior to performing cluster analysis, independence of

variables was explored by checking relation between them to ensure if there is any strong association

between variables. Number of iterations was done entering and removing different socioeconomic

variable to come to a reasonable result of classification. Finally, cluster analysis was performed using

thirteen independent variables. As the analysis produced only two automatically formed clusters, a

specific number of clusters was entered to get three classes of socioeconomic strata for this study.

Three clusters were identified that distinguish well off and poor households with middle

socioeconomic class in between. The result is summarized in Table 5-1.

The first cluster was composed of 43% of total sampled households. This cluster had the lowest

percentage of university level education attainment and all households with no formal education fell

in this cluster. There was a distinct difference in the proportion of monthly income range and

possession of letter of poor in this cluster. About 89% of households in this cluster had monthly

income in the lowest range of 700,000 IDR with about 80% of households possessing the letter of

poor. This cluster had the highest percent of households which does not have house ownership and

with temporary housing structure among three clusters. The ratio of bedrooms to total household

members was the lowest however the difference is insignificant with cluster 3. Average value of age

dependency (ratio of member from age16 - 60 to total household members) was lowest among three

clusters. Similar was the case with the mean value of ratio of employed members to total household

members. The mean value for the number of assets like bicycle, motorbike, telephone and refrigerator

owned by households in this cluster was also lowest although the difference is less between the third

cluster. From this statistics, cluster one was considered to belong to lower socioeconomic cluster

(LSEC) among three.

Cluster two was composed of 35.5% of total sample households with largest percentage (42.3%)

having university education attainment. Monthly income range was distinctively higher and

comparatively lower portion of households with the letter of poor. The result of house ownership

status was contradicting from other indicators, as cluster three had the highest percentage of

households with house ownership. However, significantly large portions of houses were with

permanent structure with better physical condition and larger bedroom ratio. It can be assumed that

living condition can be better in such house than in poor structured with ownership. It was also

supported by the result of highest portions of houses with private toilet and none of houses without

toilet. The average values for the ratio of working age members and employment ratio to the total

household members were higher than other two clusters. Also the mean values of all household assets

were large with all households having car in this cluster. This cluster presents the better living

condition and higher socioeconomic characteristics, therefore considered as higher socioeconomic

class (HSEC).

Most of the statistical results of the last cluster were in between cluster one and two which was

composed of 21.6% of total sample households. This cluster had similarities in results with other two

clusters in case of different indicators. For instance, none of the household without education and

with lowest range of income was in this cluster, which is similar to cluster two. Also the combined

percentage of higher education attainment is close. On other hand, the value for permanent structure

of houses, type of toilet, working age ratio, employment ratio, bedroom ratio and asset possessions

were similar to that of cluster one. Therefore, this cluster was considered to be in better

socioeconomic condition than the first cluster and worse than the second, so it was considered as

middle socioeconomic class (MSEC).

42

Indicators

Overall

frequency %

Cluster characteristics (Frequency %)

1(LSEC)

N=117 (42.9%)

2(HSEC)

N = 97 (35.5%)

3(MSEC)

N = 59 (21.6%)

Cate

goric

al

varia

ble

s

Education level

No education

Below senior high school

Senior high school

University

Others

0.5

19.5

41

31

8

1

35

42

22

0

0

5

34

42

19

0

12

43

37

8

Range of monthly income

Less than 700,000

700,000 – 1,400,000

1,400,000 – 2,800,000

2,800,000 – 5,600,000

Above 5,600,000

38

39

19

4

0

89

10

1

0

0

0

42

47

11

0

0

88

12

0

0

Possession of letter of poor

Yes

No

68

32

80

20

54

46

64

36

House status

Owned

Not owned

78

22

61

39

85

15

100

0

Type of house construction

Permanent

Semi permanent

Temporary

39

53

8

13

69

18

86

14

0

14

81

5

Toilet type

Public

Private

No toilet

18

77

5

27

66

7

2

98

0

27

65

8

Indicators Overall mean Cluster mean

(Higher value indicates better socioeconomic condition)

Con

tin

uou

s varia

ble

s

Household characteristics

Age dependency (16 - 60

years / total members)

Employment ratio

0.59

0.37

0.54

0.28

0.66

0.52

0.57

0.31

Housing Condition

Bedroom ratio

0.54

0.47

0.66

0.49

Asset possession

Cluster mean and % with asset possession

(Higher value indicates better socioeconomic condition)

Number of bicycles

Number of motorbikes

Number of cars

Number of telephone

Number of television

Number of refrigerator

1.01

1.24

0.07

1.38

1.15

0.55

0.79 (57%)

0.68 (56%)

0 (0%)

0.94 (50%)

0.91 (80%)

0.21 (21%)

1.32 (67%)

2.08 (93%)

0.2 (19%)

2.1 (90%)

1.52 (99%)

0.99 (76%)

0.95 (65%)

0.97 (73%)

0 (0%)

0.95 (56%)

1 (93%)

0.47 (48%)

In this way, each household was classified into three different socioeconomic strata which were used

for the subsequent analyses that follow. The proportion of each socioeconomic class for the sample

Table 5-1: Statistics of different socioeconomic clusters

[Note: LSEC = lower socioeconomic class; MSEC = middle socioeconomic class; HSEC= higher socioeconomic class;

N = number of households]


43

villages is shown in Figure 5-1. The portion of LSEC household was found to be greater (almost

48%) in village Tegalpanggung. This village had the highest population density in DIY and is located

in the core of Yogyakarta city. Employment opportunity could be one reason for the poor household

to concentrate in the city core, which avoids the travelling cost for work. Also the west boundary of

Tegalpanggung is defined by Chode River, along which informal settlement resides. Kricak had the

highest portion (40%) of HSEC households. However, difference in proportion between HSEC and

LSEC was not much (only 3.5%). Tridadi also had highest portion of LSEC households (44%)

followed by HSEC households (37%). Kricak and Tridadi are steadily growing villages in terms of

population and economic activities, as located in the periphery of Yogyakarta city. As observed

during field survey, the majority of physical living condition (housing condition) was found better in

Kricak and Tridadi with less temporary houses with bigger living area. During survey it was noted

that better off households preferred to live in urban periphery that offers bigger space, as the core

being highly dense. Also they normally owned a private vehicle, generally motor bike, so

transportation was not a big problem to reside at considerable distance from city core.

Figure 5-2 shows some of the general characteristics of different socioeconomic clusters and Figure

5-3 displays the spatial distribution of different socioeconomic classes in three villages.

Middle socioeconomic class

Lower socioeconomic class

Higher socioeconomic class

43.8

%

47.8

%

36.5

%

22.8

%

18.8

%

29.4

%

37.4

%

23.5

%

40

Figure 5-1: Proportion of socioeconomic cluster per village

(b) Education level (a) Income range

HSEC

MSEC

LSEC

HSEC

MSEC

LSEC

HSEC

MSEC

LSEC

44

Figure 5-2: General characteristics of socioeconomic clusters per village

Figure 5-3: Spatial distribution of socioeconomic classes

Tegalpanggung

Kricak

Tridadi

HSEC

MSEC

LSEC

HSEC

MSEC

LSEC

HSEC

MSEC

LSEC

(c) Housing structure

HSEC

MSEC

LSEC

HSEC

MSEC

LSEC

HSEC

MSEC

LSEC

(d) Toilet type


45

5.2 Measuring Dimensions of Access to PHC

Depending on the nature of indicators developed, each dimension of access to PHC was measured

and compared using descriptive statistics. This allowed to observe variation in state of access from

different dimensions between households from different villages, visiting public and private health

facilities and belonging to different socioeconomic classes.

5.2.1 Descriptive Statistics for Availability of PHC

For availability, questions like type of health facilities; need to get an appointment before visiting

PHC facility; waiting time to get check up by doctors and provision of medical supply were asked.

Around 75% of respondents reported to visit public healthcare facilities which were available at

every sub districts. About 18% of total respondents mentioned about the need to get an appointment

before visiting private clinics. None of the respondents stated about difficulty in getting appointment

as it could be easily done by telephone. In general availability of medicine did not seem to be a

problematic factor, as 97.5% of respondent mentioned that it was supplied by the public health

facilities itself or, if not, by a pharmacy near to the facilities. Among all these issues, respondents

expressed discontent only with the waiting time before getting check up especially in public health

clinics, puskesmas. This problem was remarkable as large percentage of respondent with 42% of total

were dissatisfied with long waiting time in health facilities. Hence, waiting time seemed to be an

important factor to be considered in availability in this study. As length of waiting time differed

depending on type of health facilities, for instance public or private healthcare facility, attributes

related to availability was studied separately for different facilities being used by respondents.

Type of primary healthcare facilities in use

Government healthcare centre, Puskesmas, was found to be the most common healthcare facility in

study area for primary healthcare. About 65% of total respondents from all three villages mentioned

puskesmas as the first PHC facility they visited. Total percentage of households visiting puskesmas

for PHC was about 53%, 79% and 60.5% for Kricak, Tegalpanggung and Tridadi respectively.

Private clinic was found to be the next common facility with 18% of total households visiting it;

mostly those belonged to higher socioeconomic class. Remaining 17% visited hospitals or sub-

puskesmas. The proportion of households visiting private clinics was greater in Kricak with about

22.5% followed by Tridadi with 20.5% and 10% in Tegalpanggung. This can be related to the result

of socioeconomic classification. As the portion of better off households was higher in Kricak, so is the

case with the visit to private clinic. Similarly, majority of households in Tegalpanggung belonged to

lower socioeconomic class so the proportion of visit to private clinics is least among three villages.

The proportion in Tridadi was in between Kricak and Tegalpanggung. However the difference is very

less with Kricak in both cases.

To check if socioeconomic status influenced the selection of PHC facility, relation between

socioeconomic class and type of facility visited was checked. Significant medium association was

found between these variables with Cramer‟s statistic .45 (p < .001). It was assumed that the

socioeconomic status influenced the selection of particular health facility to some extent. Hence the

perception on different dimensions of access based on type of facility being visited might change

accordingly.

46

Waiting time in PHC facility

Waiting time in PHC facility before getting check up by doctors showed a great range of difference

starting from 5 minutes to 3 hours. The range differed depending on the healthcare facility visited

which is shown in Table 5-2. Difference of around 50 minutes was found in the average waiting time

at puskesmas and private clinic. It was noted that the mean value for public facilities like hospitals,

puskesmas and sub puskesmas, was greater than the maximum waiting time for private clinics. The

overall mean values of waiting time for Kricak, Tegalpanggung and Tridadi are 50, 60 and 45

minutes respectively, which did not show much variation. Further subjective perception on the

waiting time was asked using a 5 point Likert scale starting from very short to very long. Figure 5-4

shows the percentage of overall and respondents‟ in village and per facility type that were

categorized in each level of Likert scale based on their response. Respondent‟s perception about

waiting time in health facilities was not very different when compared between villages. But large

difference was observed in case of private and public healthcare facilities. Subjective perception was

assumed to vary with personal as well as socioeconomic characteristics of respondents. For instance,

about 10% of respondents visiting a private clinic stated „long‟ for the waiting time of 20-30 minutes.

On other hand, about 25% of respondents visiting puskesmas stated „normal‟ for the waiting time

from 45-90 minutes.

% of

households

Waiting time in minutes

Facility visited for PHC Minimum Maximum Mean

Hospital 8.4 15 120 44

Puskesmas 65.2 10 180 65

Sub –puskesmas 8.7 10 90 38

Private Clinic 17.7 5 30 15

Overall waiting time in minutes before

doctor's check up without considering

different facility visited for PHC

100 5 180 52

Table 5-2: Descriptive statistics of waiting time in health facilities

Figure 5-4: Percentage of subjective perception on waiting time in

PHC

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Hospital Puskesmas Sub-

puskesmas

Private

clinic

Very long

Long

Normal

Short

Very short0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


puskesmas

Private

clinic

Very long

Long

Normal

Short

Very short

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Tegalpanggung Kricak Tridadi

Very long

Long

Normal

Short

Very short


47

5.2.2 Descriptive Statistics for Accessibility to PHC

Physical accessibility was measured in terms of travel impedance. Objective and subjective questions

were asked regarding travel distance, travel time and mode of transport to reach the PHC facility. For

accuracy, geographic distance from household point to the reported public health facilities was

measured in GIS instead of using reported travel distance by respondents. Overall mean for measured

travel distance was about 1450 meters and travel time was about 10 minutes. Majority of respondents

around 70% stated the travel distance to be near and very near when asked in a 5point Likert scale.

Similar responses were found about travel time to reach the PHC facility. Only about 3% of total

respondent stated far for travel distance. Remaining answers were stated as normal. Figure 5-5 shows

the perception of respondents regarding travel distance and travel time for each sample village.

Regarding the mode of transportation, large portion of respondents (50.5%) used motorbike to visit

the healthcare centre and about 39.5% used bicycle or walk. As the physical distance was not a

problem, only a small portion (7%) of respondents used public transportation and remaining 3%,

those from well off class, travelled by car. Depending on respondent‟s opinion, it can be said that in

general physical accessibility did not seem to be a major problematic factor in access to PHC in the

study area.

5.2.3 Descriptive Statistics for Affordability of PHC

Affordability was measured including direct and indirect expense of healthcare service by asking

costs and opinion about various healthcare costs. Possession of health card issued by local

government, „Katu Sehat’ was asked as it was a relevant aspect in affordability that influences

respondents‟ subjective perception. About 55% of households stated to have the card, among which

54.5% are from lower socioeconomic class. Majority of households (50%) which did not have health

card belonged to the higher socioeconomic class. Primary healthcare service cost was found to be

nominal in all puskesmas even without the card. However, it differed and was higher for private

clinics. As private clinic was visited mostly by well of class, majority of total respondents stated

service cost to be inexpensive. When asked about travel cost, 93% of respondent stated it to be

normal or inexpensive. Similarly, 80% said medicine cost to be normal or inexpensive as in most

cases medicine cost was covered by the health card. After asking about individual costs, opinion on

the total PHC cost was asked. Only 13 % of total respondents expressed it as expensive. This could

be due to medicine cost in cases when respondents did not have health card or if medicine was not

Figure 5-5: Percentage of subjective perception on travel distance and travel time to PHC

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Very far

Far

Normal

Near

Very near0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Very far

Far

Normal

Near

Very near

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Very long

Long

Normal

Short

Very short0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Very long

Long

Normal

Short

Very short

48

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


puskesmas

Private clinic

Expensive

Normal

Inexpensive

Very inexpensive

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


puskesmas

Private clinic

Expensive

Normal

Inexpensive

Very inexpensive

provided free of cost by health facilities. 34% of total respondents said that total cost to be normal

and remaining 53% said it was inexpensive. The percentage of perceived attributes on total cost of

PHC per village and healthcare facility type is presented Figure 5-6. Although in general, people

have positive to neutral opinion about healthcare cost, in some cases up to 20% of respondents

expressed dissatisfaction towards it. It may not be a major issue but deserves attention.

Figure 5-6: Percentage of subjective perception on total cost for PHC

5.2.4 Descriptive Statistics for Acceptability and Adequacy of PHC

Acceptability and adequacy were measured from respondents‟ perception on issues related to cultural

or religious preference in selecting particular healthcare facility. Factors like preference towards

gender of medical personnel, physical appearance of facility, service opening hour, trust in medical

ability and opinion about inter-personal treatment from facility personnel were investigated. When

respondents were asked if choice given, will they have any religious preference in choosing particular

healthcare facility, only 12.5% of total respondents said yes. Further a question was asked about the

gender preference of doctors in PHC facility. At the first instance, majority of respondents answered

that gender of doctor was not an issue for healthcare. Question was then elaborated giving an option,

if there were equal number of male and female doctors in health facilities would they have any

preference to be treated by particular gender of doctor. Then 42.5% of respondent mentioned if

applicable they would prefer female doctors for female patients and only 9% said male doctors to

male patients. However, even after the given option, 48.5% of respondents still stated that gender of

doctors did not matter for both male and female patients. Figure 5-7 (b) displays the percentage of

two opinions on gender preference when applicable, by considering preference of male or female

doctors to respective patients as „yes‟ and no gender preference for both as „no‟. This result showed

that religious factor was not a prominent issue while selecting a healthcare facility though people had

certain preference towards the gender of doctors especially for female if given an option. However,

these issues did not seem to significantly influence peoples‟ acceptability of healthcare facility while

evaluating overall access to the facility at present context. Higher literacy rate could be a reason for

better state of acceptability as highly significant association was found between the level of literacy

and score of acceptability index (Cramer‟s V = .513, p < .001). Also the province was homogeneous

in terms of religion as 90% of population was Islamic. Such issues related to acceptability are

relevant in areas which are more heterogeneous in terms of religion, ethnicity or race.

While measuring adequacy, respondents did not complain about physical appearance or cleanliness of

the PHC facility, both public and private, as 99.3% of total respondent were satisfied with the


49

cleanliness of healthcare facility they visited. Opening hour of facilities also did not seem to be a

major issue, as 87% of respondent stated that the opening hour suits their working schedule.

When asked about the trust in medical ability of healthcare facility visited, 93% of respondent had

positive to neutral opinion and only 7% said it was bad. All respondents who were unsatisfied visited

puskesmas for healthcare. Majority (89.5%) of those who visit private clinic stated the medical

ability to be good and remaining expressed neutral perception. Chi square test was done to see if any

relation exists between the opinion in medical ability and the facility visited. Significant medium

association was found between these variables with Cramer‟s statistic of .41(p < .001). Although the

association was not very strong, it can be said that perception on medical ability, to some extent, was

based on the type of facilities people visited.

A question was asked about the feeling of inter-personal treatment by medical personnel. About 40%

of respondents expressed dissatisfaction and answered it to be bad, out of which 80% were visiting

puskesmas. 35% of respondents stated that the personal behaviour was normal and 25% said it was

good. Chi square test between facility type and opinion of personal treatment showed highly

significant medium association between these variables similar to the previous case of trust in medical

ability and type of facility (Cramer‟s V = .3, p < .001). Perceived opinions on different factors

related to acceptability and adequacy are shown in Figure 5-7.

From interviews with households and puskesmas staffs, it was known that there are four different

hierarchy of service provision by public health facilities; VIP, 1st class, 2nd class and 3rd class

depending upon quality of service and service cost that patients are willing to pay. Priority for service

was given starting from VIP to 3rd class. Patients with health card seeking for free PHC received the

3rd class service. When asked about difference in such services, respondents did not complain about

quality of service but mentioned about unfriendly behaviour of medical staffs.

(b) Gender preference for doctors

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


No

Yes0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


No

Yes

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


No

Yes


(a) Religious importance

(d) Trust on medical ability

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Bad

Normal

Good

Very good

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Bad

Normal

Good

Very good


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Dirty

Normal

Clean

Very clean

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Dirty

Normal

Clean

Very clean


(c) Physical appearance of

PHC facility

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


No

Yes


50

Figure 5-7: Percentage of subjective perception on factors related to acceptability and adequacy

5.2.5 Overall Satisfaction Level with Access to PHC

After asking about the subjective opinion on each dimensions of access, overall satisfaction level with

access to PHC was also asked in 5 point Likert scale, considering all the factors in dimensions.

Majority of respondent with 48% of total expressed dissatisfaction or very dissatisfaction. 34% said

satisfied or very satisfied and remaining 18% expressed neutral opinion about the access to PHC

facility. The satisfaction level was then checked for different socioeconomic classes. Large portion of

households (45%) expressing satisfaction towards access were from higher socioeconomic class and

majority of households (51%) expressing dissatisfaction belonged to lower socioeconomic class.

When level of satisfaction was checked for different types of health facilities, remarkable difference

was found between public and private facilities. Around 60% of respondents visiting puskesmas

reported to be unsatisfied whereas this figure was less than 5% for private clinics. These results can

be interpreted with significant medium association between socioeconomic status and type of facility

being visited found by Chi square test in section 5.2.1. Figure 5-8 presents different level of

satisfaction within villages, socioeconomic cluster and per healthcare facilities. Significant strong

association was found between the type of facility (private and public) and the level of satisfaction to

access to PHC (Cramer‟s V = .51, p < .001).

Further to ensure the level of satisfaction, respondents were asked if their income is doubled do they

still want to visit the same PHC facility. 38% of respondent said they will visit the same facility and

62% said they want to change the facility. Chi square test was computed to see if there is any

association between the satisfaction level and willingness to change the health facility. Result showed

highly significant medium association between these attributes (Cramer‟s V = .42, p < .001). This

indicates that respondent‟s willingness to change the health facility was related to their satisfaction

level with access to healthcare facility and these attributes were related to each other.

(e) Inter-personal treatment by medical personnel

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Very bad

Bad

Normal

Good

Very good

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


Very bad

Bad

Normal

Good

Very good


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Hospital Puskesmas Sub puskesmas Private Clinic Hospital Puskesmas Sub-puskesmas Private clinic


51

Figure 5-8: Perceived satisfaction level with access to PHC

5.2.6 Influencing Factors to Overall Satisfaction Level

One of the research questions of this study was to find dominant factors that influence the evaluation

of perceived access to PHC. Answer to this question can be used to focus and prioritize such

dominant factors for further improvement of access. It was assumed that perception on those factors

had some relation with respondents‟ overall satisfaction level with access to PHC. During household

survey, overall satisfaction was asked after asking respondent‟s perception on each factor related to

dimensions of access. From earlier analysis in this section, majority of respondents expressed

dissatisfaction with waiting time in health facilities and inter-personal behavior by medical personnel.

All other factors related to dimensions of access to healthcare were found to be less problematic as

perceived by respondents. Hence, to see the relationship between factors related to each dimension of

access and the satisfaction level Spearman‟s correlation was computed between the subjective

opinion on different factors and overall satisfaction level. Aggregated 3 point Likert (satisfied;

normal; unsatisfied) was used, as it showed significant and stronger correlation between variables

than the 5 point Likert. Table 5-3 (a) shows the correlation matrix between factors under dimensions

of access and the satisfaction level when all samples were taken. Only few factors were found to have

statistically significant correlation between each other and with the final satisfaction level. Subjective

opinion on availability (waiting time in healthcare facility) and adequacy (inter-personal treatment)

showed comparatively strong positive correlation (r = .546 and r = .600 respectively at p < .001) with

the satisfaction level. Perception on medical ability of medical personnel also showed a significant


(a) Village

HSEC MSEC LSEC

(b) Socioeconomic class (c) Healthcare facilities

Private

clinic


Puskesmas

52

relation with the overall satisfaction level, however with low correlation coefficient (r = .232,

p<.001).

Similar correlation was repeated for households visiting to different healthcare facilities to see if

other factors have any relation with the satisfaction level depending on the choice of facility type. For

public facility like puskesmas, almost similar result was found as earlier, which is shown in Table 5-3

(b). Satisfaction level of respondents was found to be influenced only by waiting time and inter-

personal treatment. There was no strong correlation between other factors except for travel distance

and travel time. Hence, waiting time and personal treatment was found to be dominant factors in

determining peoples‟ satisfaction with access to PHC in case of public health facilities.

Correlation computed for private clinics displayed different results than previous ones. Correlation

matrix for private clinics is shown in Table 5-3 (c). In case of private clinics, respondent‟s overall

satisfaction level showed significant positive correlation with their perception on other factors like

travel distance, cost and medical ability. Like in previous cases, satisfaction with waiting time and

personal treatment also had significant positive relation with the overall satisfaction.

Also, there was significant relation between factors like waiting time, medical ability and personal

treatment which was missing in case of puskesmas. As private clinics did not have provision of free

healthcare service unlike other public facilities, affordability was an important factor that influences

peoples‟ satisfaction level to some extent. However, the correlation was not very strong, as majority

of households visiting private clinics belonged to higher socioeconomic class who expressed the

service cost to be normal.

SO LOS WT TD TT TC CI GP PT PA MA

LOS 1

WT .546** 1

TD -.090 -.014 1

TT -.030 .048 .610** 1

TC -.018 .002 .136* .078 1

CI .106 .113 .093 .033 -.037 1

GP .008 .071 .043 .026 .050 .130* 1

PT .600** .225** -.126* -.068 -.032 .113 .021 1

PA .067 .050 .171** .117 .052 .050 .076 .015 1

MA .232** .067 -.080 -.038 -.173** -.011 -.103 .251** -.002 1

**. Correlation is significant at the 0.01 level (2-tailed)

*. Correlation is significant at the 0.05 level (2-tailed)

SO = subjective opinion, LOS = level of satisfaction with overall access to PHC, WT = waiting time in PHC facility, TD =

travel distance to PHC facility, TT = travel time to reach PHC facility, TC = total cost of PHC, CI = cultural importance in

choosing PHC facility, GP = gender preference of doctors, PT = personal treatment in PHC facility, PA = physical

appearance of PHC facility, MA = trust on medical ability of PHC facility

(a) Correlation matrix for total sample households


53


LOS 1

WT .501** 1

TD -.116 -.006 1

TT -.008 .111 .587** 1

TC .025 .058 .054 .016 1

CI .086 .128 .079 .046 -.092 1

GP -.062 .045 .068 -.022 .071 .156* 1

PT .552** .092 -.090 -.014 .047 .103 .042 1

PA .119 .113 .159* .210** .061 -.028 .133 .071 1

MA .046 -.071 -.038 -.040 -.100 -.014 -.168* .165* -.032 1



(b) Correlation matrix for samples visiting puskesmas


LOS 1

WT .695** 1

TD .411** .260 1

TT .019 -.126 .338* 1

TC .309* .292* .180 .080 1

CI .269 .215 -.007 -.157 .026 1

GP .241 .164 .154 -.012 .009 .258 1

PT .432** .367* .040 -.092 -.050 .387** .110 1

PA .110 .248 .135 -.091 .059 .312* -.013 .098 1

MA .529** .462** .368** .044 -.018 .380** .236 .290* .201 1



(c) Correlation matrix for sample visiting private clinics

Table 5-3: Correlation matrices between perception on factors of access and overall satisfaction

level with access to PHC

5.3 Synthesis of Indicators to Develop Summary Scores

Perceived satisfaction level of respondents across indicators was used to develop summary scores for

all dimensions. A total of 15 indicators measured in 5 point Likert scale were used in this process

which is shown in Figure 5-9. As acceptability and adequacy lacks objective indicators, subjective

perception about different factors related to access was used. While developing such summary scores

for all dimensions, equal weights were applied to each indicator. Distributions of summary scores for

each dimension are demonstrated by using box plot in Figure E-1 in Appendix E.

For easier interpretation, indicator values were standardized with ratio scale properties (standard

score, i = raw score i / maximum raw score). Though earlier analysis in section 5.2, showed

54

significant variation in the perception of respondents between indicators under same dimension, equal

weights were used for all indicators, based on the equal importance of all five dimensions in the

concept of access to healthcare. For instance, out of three indicators under availability, only waiting

time in health facilities was found to be the problematic issue in context of study area. However, ease

of getting an appointment to visit health facilities and provision of medicine supply cannot be less

prioritized while evaluating the state of availability related to overall access to healthcare. All these

factors might hold equal importance in evaluation of access at wider perspective of policy making.

Figure 5-9: Synthesising indicators to develop summary scores for dimensions of access to PHC

Perceived state of access

to PHC viewed from

different dimensions

Perceived satisfaction

level with:

Accessibility

0.5

0.5

- Travel distance to PHC

- Travel time to reach PHC

Availability

-Ease to get appointment

-Medicine availability

- Waiting time in PHC

0.33

0.34

0.33

Affordability

-Travel cost to reach PHC

-Service cost of PHC

- Medicine cost

-Total cost of PHC

0.25

0.25

0.25

0.25

Acceptability

-Religious aspect

-Gender preference of

medical personnel

0.5

0.5

Adequacy

-PHC service opening hour

-Physical appearance of

PHC facility

-Medical ability of medical

staffs

-Inter-personal treatment

from medical staffs

0.25

0.25

0.25

0.25


55

Average summary score was calculated by dividing the sum of total scores with the number of cases

for each villages, health facilities and socioeconomic classes. This score was used as the

representative value for respective cases. Average summary scores of dimensions for villages and

healthcare facilities are plotted in a radar chart shown in Figure 5-10. These scores are listed in Table

E-1 in Appendix E.

Accessibility scores are highest for villages Tegalpanggung and Kricak as compared to other

dimensions within villages. As Tridadi belongs to rural regency, geographic distance was

comparatively larger than other two villages. Also considerable number of respondents from Tridadi

reported to visit health facility in City of Yogyakarta rather than near facilities within their sub

district. Affordability and acceptability also scored higher values as compared to availability and

adequacy in all three villages. These scores somehow refers to the result obtained in section 5.2 where

only waiting time in health facilities and inter-personal treatment from medical personnel were found

to be a concern of respondents‟ dissatisfaction. There was small difference in scores of each

dimension between villages, which indicates similarity in state of access in villages. However

satisfaction level of respondents on different dimensions of access within each village showed

remarkable variation, for instance difference in accessibility score with availability and adequacy.

Remarkable variation can be seen in the scores when different types of health facilities used by

respondents were compared. Figure 5-10 (b) gives a clear interpretation of how perceived satisfaction

level of respondents varied based on the type of health facilities they visited. The score of availability

and adequacy are significantly higher for private clinics as compared to other public facilities. As the

satisfaction level with inter-personal treatment and the quality of service of private clinics was

significantly higher than that of other public facilities. And these scores are lowest for puskesmas. But

in case of affordability, there is swap in the scores. It is very low for private clinics when compared

between scores of other dimensions within same facility. Also the score is lowest when compared with

affordability scores of other facilities, whereas it is much higher for puskesmas and sub puskesmas.

The reason behind can be related with the findings of section 5.2. As public health facilities provides

free primary healthcare and medicine to patients with health card and majority of patients visiting

puskesmas, for example, have the health card. Hence, respondents visiting public health facilities

expressed higher satisfaction towards affordability. Although majority of respondents visiting private

clinics were from higher socioeconomic class, they reported the cost of service and medicine to be

relatively expensive. As the service is free or in nominal cost, majority of population visits

puskesmas. This results in a long waiting queue for check up in the facility unlike in private clinics.

Sub puskesmas has the highest score for accessibility and that of hospital is lowest. These figures well

fit the actual context as sub puskesmas are the supporting health facilities allocated per sub districts.

There are normally two to three sub puskesmas per sub districts, whereas hospitals are the biggest

health facilities in terms of service capacity and provision and are allocated per regencies. So the

difference in geographic distance to these facilities was large. Acceptability had relatively higher

scores with small variation in both cases of villages and facilities. As respondents expressed that

religious factors and gender issues were not relevant matters of concern in choosing health facilities

and they were satisfied with the existing situation in terms of acceptability.

56

Figure 5-11 shows the standardized residuals of percentage of respondents‟ perceived satisfaction

with each indicator of access. Positive residuals imply that frequency of satisfied respondent is higher

than expected average and negative residual means the reverse in all cases.

Frequency of respondents, satisfied with both indicators of accessibility was lower from village

Tridadi as compares to other two, showed in Figure 5-11 (a). In availability, frequency of satisfaction

with waiting time in facility was distinctly low in Tegalpanggung, whereas it was higher for Kricak.

Satisfaction with indicators of affordability was higher in Tegalpanggung. Tridadi had low

satisfaction with travel cost which can be seen in relation with lower satisfaction with travel distance

and time. Difference was less for indicators of acceptability, as respondents‟ perception was more

homogeneous in these factors. Variation between villages was observed in case of adequacy.

Comparatively Tridadi was in better satisfaction level with medical ability and inter-personal

treatment of medical personnel. And, inter-personal treatment was a notable problem with

respondents in Tegalpanggung.

Hospital

Puskesmas

Sub puskesmas

Private clinic

(b) Healthcare facilities

(a) Villages

Tegalpanggung

Kricak

Tridadi

Figure 5-10: Summary score chart for dimensions of access to PHC

For: (a) sample villages and (b) different healthcare facilities


57

More distinct variation was observed in case of different health facilities displayed in Figure 5-11

(b). It was clear that when compared, large portion of respondents expressing problem with physical

distance were the one visiting hospitals. Significant difference was seen in case of waiting time in

facility; healthcare service cost; total cost; trust in medical ability and personal behaviour between

puskesmas and private clinics. Among these indicators, only healthcare cost was the matter of less

satisfaction to respondents visiting private clinics when compared with the cost of puskesmas.

Satisfaction with waiting time, and indicators under adequacy was much higher in case of private

clinics and was low for puskesmas.

Average summary scores were used to see variation in dimensions of access across different

socioeconomic classes to observe homogeneity within villages in terms of access to PHC. Table 5-4

presents the average summary scores of dimensions for each socioeconomic class per village. There

was no distinct variation in individual dimension scores between socioeconomic classes in all three

villages. However, some variation was found between scores of different dimensions within each

socioeconomic class in Tegalpanggung and Kricak. All socioeconomic classes in these two villages

Figure 5-11: Standardized residuals for perceived satisfaction with various factors of access

Perceived satisfaction with:

1.a = Travel distance to PHC

1.b = Travel time to reach PHC

2.a = Ease to get appointment

2.b = Medicine availability

2.c = Waiting time in health

facility

3.a = Travel cost

3.b = Service cost

3.c = Medicine cost

3.d = Total cost

4.a = Religious aspect

4.b = Gender preference of

medical personnel

5.a = Service opening hour

5.b = Physical appearance of

facility

5.c = Trust in medical ability

5.d = Inter-personal treatment

(a) Villages

(b) Health facilities

58

had higher score for accessibility and lower for adequacy. Scores for affordability and acceptability

were comparatively high than availability and adequacy with small variation. Higher socioeconomic

classes in all villages had slightly higher scores for adequacy and availability than middle and lower

socioeconomic classes, whereas it had lower scores in affordability. This can be related with the type

of facility visited by different socioeconomic class, as large percentage of respondents visiting private

clinics falls in higher socioeconomic class as discussed in section 5.2.1. Tridadi was most

homogeneous in terms of variation in dimensions of access to PHC within and also between all

socioeconomic classes. As very small variation was obtained between scores of different

socioeconomic classes, radar chart could not display clear interpretation, hence it was not used.

Accessibility Availability Affordability Acceptability Adequacy

Tegalpanggung

HSEC

MSEC

LSEC

0.81

0.82

0.79

0.73

0.70

0.71

0.74

0.79

0.78

0.77

0.76

0.76

0.71

0.68

0.66

Kricak

HSEC

MSEC

LSEC

0.81

0.82

0.8

0.77

0.71

0.73

0.70

0.78

0.81

0.78

0.76

0.75

0.70

0.69

0.67

Tridadi

HSEC

MSEC

LSEC

0.72

0.73

0.73

0.76

0.70

0.71

0.74

0.76

0.75

0.76

0.77

0.76

0.75

0.72

0.70

Table 5-4: Summary score for dimensions of access per socioeconomic class

The reason for this homogeneity was due to the fact that majority of households (74%) from all

villages visited public health facilities like hospitals and puskesmas for primary healthcare. People

opt for public facilities, primarily puskesmas, because of short geographic distance; ease in medicine

availability and satisfactory quality of service in terms of medical ability and cleanliness, in addition

to nominal service cost. Hence there was similarity in perception over factors related to access from

respondents regardless of their socioeconomic status. As the number of respondents from high

socioeconomic class visiting private clinics was low (18%) as compared to the number visiting

puskesmas from same socioeconomic class, their perception was over shaded by the large number of

respondents visiting public facilities while taking average score. Variation observed for different

health facilities in Figure 5-10 (b) can also explain the variation between socioeconomic classes to

some extent. As around 80% of respondents visiting private clinics are from higher socioeconomic

class and only 19% from middle socioeconomic class, the summary scores of private clinics also

reflects the status of access for higher socioeconomic class. In this manner, it can be said that some

variation do exists between different socioeconomic classes.

5.4 Existing Situation of Access to PHC from Policy Perspective

To address the research questions of second sub objective, access to health care was analyzed as per

the official norms mentioned in existing health policy in the province of Yogyakarta (DIY). The

health policy in DIY primarily focuses on physical accessibility, availability of public health facilities

and affordability for healthcare service. Information on geographic distances to public health

facilities and total number of doctors per village, obtained from census 2005 was used for analyses.

Source of all health policies and standards used in this study was Statistics of Indonesia obtained from

Note: High values indicate high level of access; 1 = maximum value

[HSEC = Higher socioeconomic class; MSEC = Middle socioeconomic class;

LSEC = Lower socioeconomic class]


59

the Gadjah Mada University (Pustral UGM). As per the norm, there should be at least one puskesmas

headed by a medical doctor at every sub district level supported by 2 to 3 sub puskesmas headed by

medical nurses. This is related with accessibility as well as availability. It addresses the issue of

geographic distance to healthcare considering administrative boundary and availability by

considering puskesmas to population ratio per sub district. This norm was well met at existing

situation, as all 78 sub districts in DIY contained at least one puskesmas and supporting sub

puskesmas. Considering geographical distance, 70.5% of total population in the province (271

villages out of 438) was within 2 kilometres and 88% population (327 villages) within 4 kilometres,

assuming 4 km per hour as normal walking distance. All 45 villages in the city of Yogyakarta were

within the distance of 2 km from nearest puskesmas in respective sub districts covering 15.5% of total

population in the province. In terms of distance to hospital, 44.5% of total population (138 villages),

including all in city of Yogyakarta, in the province was within 4 km distance from hospital. Variation

in geographic distance to puskesmas and hospitals between villages in DIY is shown in Figure 5-12

(a) and (b) respectively. This can be related with the finding of previous analysis for accessibility in

section 5.2.2, where majority of respondents in sample villages reported physical accessibility as less

problematic issues in access to healthcare. Regency Gunung Kidul had the highest area and lowest

population density among all five in the province. Also most of villages, except those located in

central area in this regency had primarily agricultural landuse. As settlement was concentrated in the

central part of the regency, health facilities were also located accordingly. Hence, looking at Figure

5-12 (b), large number of villages in regency Gunung Kidul showed higher travel distance to hospital

except for villages located in central part of the regency. Villages at larger geographic distance from

hospitals can be related with lower population density in Figure3-2 and vice verse.

According to the health policy in DIY, ratio of puskesmas to population was 1: 120,000 and that for

hospital was 1:240,000. The ratio of puskesmas to population in all sub districts was distinctly

smaller than the standard threshold as none of the sub district had population this large. Largest

population of sub district in DIY was 117,130 in Sleman regency. Hospital to population ratio per

regency was also found smaller than the standard, except for regency Gunung Kidul which was

around 1: 250,000. Further to elaborate the analysis of availability, population to doctor ratio was

checked per sub districts using the number of doctors working at public health facilities and

population data per village in census. Analysis was done at sub district level as it was the smallest

administrative unit for the allocation of puskesmas according to health policy. Dash (2000) stated

that average doctor population ratio in Indonesia was around 1:7,000 according to the World

development report. Hence due to lack of specific standard threshold for doctor population ratio, this

ratio was taken as the benchmark for the analysis in DIY. Variation in ratio between sub-districts is

shown in Figure 5-13 (a). Out of 78 sub districts 43 had the ratio below 1: 7,000 which includes all

sub districts within city of Yogyakarta. 14 sub districts had the ratio 1:7,000 to 1:14,000 and other

14 sub districts had the ratio larger than 1: 14,000. Majority of sub districts with higher ratio was in

regency Gunung Kidul and Kulon Progo. All three sub districts containing sample villages had the

ratio below 1:7,000. Figure 5-13 (b) shows the relative variation in population to doctor ratio within

the province of Yogyakarta using standard deviation. Average ratio in the province was found to be

1:9900, with standard deviation of 12600, which was larger than the average ratio in Indonesia. Huge

variation was observed in the population doctor ratio starting from 1: 425 in regency Sleman going up

to 1: 55,000 in Gunung Kidul. Comparatively city of Yogyakarta and regency Sleman were in better

situation in terms of availability of medical doctors. This result was contradicting with the higher

dissatisfaction of respondents with long waiting time in public health facilities reported during

household survey. There was a huge difference in population doctor ratio when compared with the

average ratio of neighbouring country like Singapore.

60

Dash (2000) stated that population to doctor ratio in Singapore was 1:900. Further comparing it with

developed countries like Japan (1:600) and United State (1:400), the average population-doctor ratio

in Indonesia was found to be very large. Also when referred to the estimated ratio of physicians to

population by human resources for health, World Health Organization (WHO), the existing ratio was

found to be very large. WHO stated that although there is no fix standard of ratio of population to

medical service personnel, countries with less than 25 medical staffs including physicians and nurses

per 10,000 populations cannot achieve adequate service coverage for primary healthcare. ESCAP

(2009) stated the estimated ratio of 1 physician to 1000 population in context of Asia in WHO

statistical information system 2008. When compared with this figure, only 4 sub districts; 2 in Sleman

and 2 in City of Yogyakarta, were found to have meet the ratio estimated. This could be related to

the respondents‟ complaints about long waiting time in health facilities.

(a) (b)

Distance to puskesmas (km) Distance to hospital (km)

(a) (b)

Figure 5-12: Geographic distance to health facilities per village in DIY,

Data source: Census 2005

Figure 5-13: Population-doctor ratio per sub districts in DIY Data source: Census 2005


61

One of the major health policies in the province of Yogyakarta focuses in the provision of nominal

primary healthcare service in almost all public health facilities. As explained earlier in section 3.2.2,

better service quality, accessibility and affordability are the part of „Healthy Indonesia 2010‟ which

is one of the policy visions in line. The policy concerned to affordability was primarily focused to

facilitate poor population by providing government health insurance card, commonly known as a

health card. This policy can be considered to be well implemented as more than 55% of sample

households in Tegalpanggung and Tridadi were getting free primary healthcare service including

medicine provision with the possession of health card. This figure was around 45% for Kricak.

Among total sample, 80% of households having a letter of poor reported to have a health card, which

was a criteria for getting the health card according to the norm. Even without health card, service cost

for primary healthcare in all public health facilities was found to be nominal as 87% of total

respondents were satisfied with the total cost of healthcare. Although issue of service quality

appeared in health policy, there was not any standard threshold to evaluate the existing quality of

health service. Hence it was done based on respondents‟ perception from household survey. Quality

of service basically means peoples‟ trust on the medical ability of health facilities. When asked, only

7% of respondents expressed dissatisfaction with the medical ability in terms of quality of service. As

medical service is given by or under supervision of professional medical doctors at puskesmas, people

expressed their trust over the service quality. Other issues related to the service quality like

cleanliness of facilities, availability of equipment and laboratory service for primary healthcare were

also found to be considerably in satisfactory level.

5.5 Scale Effect in Analyzing Socioeconomic Attributes and Access to PHC

Common variables between census and primary data were selected considering their availability and

relation to socioeconomic characteristics and access to healthcare for comparative analyses between

aggregated and disaggregated data. These variables are explained in Table 5-5.

5.5.1 Comparing Individual Variables from Census and Primary data

As a preliminary analysis, information on individual variables from census 2005 was compared with

that from primary household data. Significant difference was found in the information about

socioeconomic characteristics like possession of letter of poor and health card. These variables were

important as they reflect socioeconomic status of people and also from affordability aspect of access

to primary healthcare. Census presented very low percentage (less than 2%) of households with the

letter of poor in all three villages. This figure significantly differed from the finding of primary data,

which showed around 55% of households with the letter of poor in Kricak and more than 70% in

Tegalpanggung and Tridadi. Similarly, according to census, public piped water supply was the

dominant source of water in Kricak and Tridadi. But from field survey, it was found that around 60%

households in both villages were using well water, mostly shared community well. Regarding type of

toilet in all villages, census information was a good representation, as more than 70% of households

have private toilet which was mentioned as the dominant toilet type in census. Observation over the

percentage of households possessing health card for primary healthcare also showed large difference

of 72% in Tegalpanggung and 45% in Tridadi, as census present low figures for this attribute. But in

case of Kricak, census presented higher portion of 89.5% households with the health card, which was

found to be 67.5% from household data. In contrary, census showed low portion of households with

telephone connection, 14.5% in Kricak and less than 5% in other two villages. Collected primary

data showed more than 55% of households with telephone connection in all three villages.

Aggregated data on access to healthcare was available for accessibility in terms of geographic

distance and perception on ease to reach healthcare facilities. Hence those attributes were selected for

62

comparative analysis. Variation was found in physical accessibility to healthcare when check at

different level of administrative unit. Measured distance between sample households and healthcare

facilities (hospital and puskesmas) being visited by respective households was used to compare with

census figure. Distance to hospital obtained from primary source was found to be greater in

Tegalpanggung and Tridadi. Difference was much higher for Tegalpanggung. In case of Kricak,

census showed a smaller distance than the average value of primary data. Distance to puskesmas from

villages also showed some variation.

Table 5-5: Common variables in census 2005 and household data

Table 5-6 presents the descriptive statistics of physical distances to these healthcare facilities

obtained from primary data. Although there was not large difference in the average value of measured

distance and census, except in case of distance to hospital from Tegalpanggung, large variation in

minimum and maximum range of distance was observed in most of the cases. Around 40% of

households in Tegalpanggung and Tridadi did not visit hospital for primary care. Remarkable

variation in census figure and primary data was found in Tegalpanggung. Distance to hospital from

this village was 500 meters in census, whereas none of the sample household was within this distance.

The minimum distance from primary source was 1000 m and 15% of sample households are travelling

3000 to 4000 meters to visit a hospital. Also in Tridadi, only 24% of sample was within the distance

mentioned in census (4000 m) and about 31% was between a distances of 6000 to 11000 meters.

Similarly, the distance to puskesmas in census was not found to be a good representative for

Tegalpanggung and Kricak, as large percentage of sample was travelling greater distance than as

mentioned in census. This analysis showed the inconsistency in the results obtained from primary data

source, compared to the aggregated census figures. On the other hand, census figure was found to be a

good representative of distance to hospital in case of Kricak and distance to puskesmas for Tridadi as

more than 60% of sample fell within the distance as mentioned. Figure 5-14 shows the measured

Euclidean distance from each sample households to the visited hospitals and puskesmas as reported.

This displays the variation in travel distance to health facilities within each village, prominently for

hospitals from rural village Tridadi.

Attributes related to

socioeconomic status

Description according to Census

Letter of poor

Percentage of households within village possessing the letter of poor (see

section 3.2.2 for description)

Toilet type

Dominant toilet type in village (shared public toilet; private toilet and no toilet)

being used by majority of household in the village

Water source

Dominant source of water supply in village. (River water, ground water – hand

pump, well and public piped water supply)

Telephone connection Percentage of households within village with telephone connection

Attributes related to

access to healthcare

Distance to hospital Average physical distance from village centre to nearest hospitals

Opinion about distance to

hospital

Subjective opinion about ease to reach the nearest hospital based on physical

distance in four scale; very easy, easy, difficult and very difficult

Distance to puskesmas

Average physical distance from village centre to nearest puskesmas usually

located in same sub-district

Opinion about distance to

puskesmas

Subjective opinion about ease to reach the nearest puskesmas based on physical

distance in four scale; very easy, easy, difficult and very difficult

Possession of health card Percentage of households possessing government health card


63

Based on the nearest geographic distance, census stated a subjective opinion on the ease in reaching

those health facilities. As distance was not considered to be a problem in case of these villages, all of

them were given „easy‟ or „very easy‟ status in terms of physical accessibility in census. The

perception over physical distance obtained from primary data also stated accessibility to be less

problematic, in spite of the fact that people does not necessarily visit the nearest healthcare facilities

as per assumption. However some variation exists in the perception. Chi square test was done to see if

there is any association between the respondent‟s perception over accessibility measures and the

mode of transport they use to reach health facilities, which was asked during survey. Significant

medium relation was found between these attributes with Cramer‟s statistics of .46 (p < .05). Hence,

there was large variation in the respondents‟ perception over travel distance.

Due to lack of additional information on other dimensions of access to healthcare in census data, this

study was limited to compare variables related to accessibility, although it did not seem to be a

problematic issue in access to PHC in study area.

Figure 5-14: Euclidean distance from households to heath facilities

Tridadi

Kricak

Tegalpanggung

Tridadi

Kricak

Tegalpanggu

ng

64

Comparing census distance to hospitals with primary data (all distance is in meters)


Distance in census = 500 m Distance in census = 3000 m Distance in census = 4000 m

Primary data

% of households within distance

Less or equal to 1000 = 0% Less or equal to 3000 = 64% Less or equal to 4000 = 24%

1001 - 2000 = 35% 3001 - 6000 = 0% 4001- 6000 = 5%

2001 -3000 = 9% 6001 -7000 = 3% 6001 -7000 = 15%

3001 - 4000 = 15% greater than 7000 = 16%

Do not visit hospital = 41% Do not visit hospital = 33% Do not visit hospital = 40%

Mean = 2160 Mean = 2255 Mean = 5215

Minimum = 1000 Minimum = 1225 Minimum = 1050

Maximum = 4045 Maximum = 7200 Maximum = 11000

Comparing census distance to puskesmas with primary data

Distance in census = 500 m Distance in census = 500 m Distance in census = 3000 m

Primary data

% of households with distance

Less or equal to 500 = 18% Less or equal to 500 = 30% Less or equal to 3000 = 60%

501 – 1000 = 44% 501 – 1000 = 53% 3001 – 6000 = 3%

Greater than 1000 = 17% 1001 – 3000 = 7% Greater than 6000 = 7%

Do not visit puskesmas = 21% Do not visit puskesmas = 28% Do not visit puskesmas = 30%

Mean = 750 Mean = 625 Mean = 1970

Minimum = 240 Minimum = 1210 Minimum = 500

Maximum = 1210 Maximum = 3000 Maximum = 7850

Table 5-6: Distance to healthcare facilities from Census and primary data

5.5.2 Comparing Variation within Village with Variation at Sub district Level

Coefficient of variation (CV) was computed to further compare variability in selected socioeconomic

and access attributes at village level with that at sub districts level, using primary and secondary data

respectively. Variation in common socioeconomic attributes was found higher for all villages as

compared to their sub districts. This is shown in Table 5-7 (a). Although, private toilet was the

dominant toilet type in all sample villages, some variation was found within each village, whereas

census had a single value of toilet type for all three villages. Significant variation was observed in the

possession of health card in village Kricak and Tridadi which is directly related to affordability in

access to PHC.

Variation, in terms of access to healthcare, was found in physical distance to hospitals and puskesmas

at village level in Tegalpanggung, shown in Table 5-7 (b). This variation could not be observed at its

sub district level, as all villages in this sub district were given same value for respective distance

according to census. The variation in physical distance to puskesmas was significantly large in

Tridadi when compared with the variation at its sub district using aggregated village data. As Tridadi

belongs to rural regency, significant portion of respondent reported to visit health facility in

Yogyakarta city, resulting in larger variation in travel distance as well as their perception over it. In

case of Kricak, variation was found to be less than the variation in its sub district, however the

difference was small. Regarding the relative subjective perception on these distances, sub districts of


65

Tegalpanggung and Kricak does not show any variation as all the villages within these sub districts

were stated as having a „very easy‟ accessibility to mentioned facilities in census.

Table below displays the variation at sub district level using census data available for each village

within the sub districts. Rows specified as „census(s)‟ shows this variation. Rows specified as

„Primary data (v)‟ displays the variation within sample villages using primary household data.

Purpose is to compare variation within village to that with its corresponding sub district.

Tegalanggung belonged to sub district Danurejan; Kricak to Tegalrajo and Tridadi to Sleman, hence

comparison is done correspondingly.

(a) Socioeconomic variability in terms of coefficient variation (CV)

Data source

Sub-district (s)

/ Village (v)

Type of

toilet

Water

source

Telephone

connection

Possession of

letter of poor

Possession of

health card

Census (s) Danurejan NA NA 0.72 0.45 0.54

Primary data (v) Tegalpanggung 0.28 0.60 1.03 0.60 0.70

Census (s) Tegalrajo NA 0.16 1.18 0.39 0.13

Primary data (v) Kricak 0.24 0.54 1.50 0.86 1.12

Census (s) Sleman NA 0.16 0.8 0.64 0.22

Primary data (v) Tridadi 0.22 0.48 1.01 0.65 0.96

(b) Variability in accessibility to PHC in terms of coefficient variation (CV)

Data source

Sub-district (s) /

Village (v)

Distance

to hospital

Subjective

perception on

distance to hospital

Distance to

puskesmas

Subjective

perception on

distance to puskesmas

Census (s) Danurejan NA NA NA NA

Primary data (v) Tegalpanggung 0.435 0.276 0.375 0.279

Census (s) Tegalrajo 0.556 NA 0.707 NA

Primary data (v) Kricak 0.450 0.264 0.675 0.288

Census (s) Sleman 0.520 0.248 0.476 0.248

Primary data (v) Tridadi 0.530 0.261 0.941 0.257

Table 5-7: Socioeconomic and access to PHC variability in terms of CV

[NA = not applicable; if census have one value for all villages within that sub district]

66

6. Discussions on Findings

This chapter presents the analytical discussion of the results obtained to address sub objectives of

this study in three main sections. Main findings and limitations in methodologies applied in this

research are discussed. In the first section, results of variation in access between villages and

different socioeconomic classes are discussed. Perceived major factors that influences the overall

satisfaction level are also discussed in this section. Second section discusses the result of access to

healthcare in relation to contextual heath policies comparing it with results at micro level. Last

section contains a discussion over the findings of variation in results obtained from aggregated

and disaggregated data source.

6.1 Sub-objective 1: Measuring Access to PHC at Micro Level

Socioeconomic stratification

To address one of the research questions of sub objective 1, related to variation in access across

different socioeconomic strata, sampled households were classified into three classes using two step

cluster analysis. This stratification was relevant to measure intra village variation in terms of

socioeconomic characteristics which was assumed to have direct relation with access to healthcare.

Analytical discussion over the strength and limitations of the method used for stratification is

important as large part of subsequent analyses and findings was related to this result. Along with

advantages offered by cluster analysis explained in section 4.6.1, it had some limitations which were

experienced in this study. As explained by Wong (2006), cluster analysis require detailed and often

debatable operational decisions throughout statistical procedure. First, measurement of some form of

similarity in attributes between samples is needed in order to decide number of clusters. Once

different clusters were obtained, next step was the profiling of clusters to give them appropriate

labelling based on the composition of various input attributes. Although clusters were labelled

according to the characteristics exhibited by them, it was somehow a subjective judgement. For

instance, middle socioeconomic cluster in this study did not exhibit terribly distinctive characteristic,

rather it had some common characteristics from both higher and lower socioeconomic clusters. This

could be one reason that summary scores for dimensions of access developed for each socioeconomic

class within 3 sample villages in section 5.3 (Table 5-4) did not show much variation. Also other

statistical analyses, for example, Chi square did not have strong association (Cramer‟s V = .45)

between these classes and other variables like selecting type of health facilities. Another limitation is

that no ranking of individual cases within each cluster can be obtained as the samples were either in

or out of the cluster. Other cluster analysis such as K mean offers relative ranking of cases within

each group based on how strongly cases represent the belonging cluster, this was lacking in two step

cluster analysis. However, due to the advantage of simplicity and mainly its ability to create clusters

using both categorical and continuous variable two step cluster analysis was used for this study.

One of the assumptions made for selection of sample village was socioeconomic heterogeneity within

villages. This assumption was met as each sample village was composed of considerable percentage

of high and low socioeconomic classes. Also the assumption that poor informal settlement exists along

river bank, was met in case of village Tegalpanggung (see Figure 5-3).

Measuring access to PHC

Factors related to physical accessibility and availability of health facilities has been important

concerns in most of the empirical studies in evaluating access to healthcare. Literatures like Black,


67

Ebener et al. (2004); Bagheri, Benwell et al. (2005); Amer (2007) etc have been able to demonstrate

significant variation in these spatial component of access while evaluating health service variation.

However these dimensions in general, were not found to be the major problematic issues in this study

as the result of high priority given in the implementation of existing health policies related to physical

accessibility and availability of public healthcare facilities. Below 10% of total respondents,

dominantly from rural regency Sleman, stated physical distance to hospitals or puskesmas to be far.

This was not due to absence of health facilities within their administrative boundary, rather due to

user‟s personal preference to visit facilities in city of Yogyakarta.

Along with above mentioned spatial components, affordability and acceptability has also gained

priorities in evaluating equitable access to healthcare referring to literatures like Andersen,

McCutcheon et al. (1983); Fosu (1989); Obrist, Iteba et al. (2007) etc. However, as the result of free

or nominal cost of primary healthcare service including medicine supply in public facilities,

affordability in general was perceived as less problematic issue in the study area. Only about 13% of

total respondents were discontent with the total cost for primary healthcare. Noticeable fact related to

affordability was that none of the respondents from low socioeconomic class visited private clinic in

spite of their feeling about better service quality, short waiting time and good inter personal treatment

being provided by private clinics. The reason was directly linked with the higher service cost in

private facilities. This was confirmed when majority of respondents visiting public hospitals or

puskesmas stated that they would like to visit private facilities for primary healthcare, if their income

gets doubled. Also the satisfaction level with total cost showed a negative correlation with the

satisfaction level with trust in medical ability of health facilities (see Table 5-3. a). As respondents

visiting private clinics stated the cost to be higher, however they were very satisfied with medical

ability of the facility. Therefore affordability still had significant relevance in access in this manner.

Long waiting time under availability and unfriendly inter-personal treatment by medical staffs under

adequacy, were found to be dominant factors influencing the overall satisfaction of respondents

especially visiting public health facilities. Only satisfaction with these two factors showed highly

significant positive correlation with the overall satisfaction level (r = .546; r = .600 respectively at p

< .001). All other factors under five dimensions of access did not seem to influence the overall

satisfaction level of public facility users. Hence, it can be assumed that people tend to emphasize on

negative issues when asked about their satisfaction with present situation of access to PHC.

Satisfaction with factors like travel distance, service cost and trust in medical ability were important

in case of private health facility users, as these perceptions were found to influence the overall

satisfaction of people going to private clinics. Noticeable difference was observed in the perception

about similar waiting time between public and private facility users. As puskesmas provided free

primary healthcare for poor households, people might have accepted the longer waiting time or they

might be used to this situation hence, found it normal in general case. But majority of households

visiting private clinics belonged to higher socioeconomic class, who might be willing to pay more but

expects faster service. Due to the hierarchy in service provision by public health facilities, patients

seeking for free service might have experienced unfriendly behaviour by medical personnel. This

might create an issue of inequality in terms of inter-personal treatment among different

socioeconomic strata.

To evaluate combined effect of relevant factors under each dimension of access, summary scores were

developed which showed some limitations in this study. Along with the advantage of easy

interpretation especially at decision making level, a common disadvantage of developing summary

scores was that it loses underlying information to some extent. For instance, the problem of waiting

time and inter-personal behaviour in public health facilities was over shaded by the higher score of

other corresponding indicators while developing availability and adequacy summary scores. Another

68

limitation of this process was the weighting system. Although equal weights were applied to each

indicator, the relative value differed due to different numbers of indicators under different

dimensions. Larger the number of indicators within a dimension, lesser will be the impact while

developing a summary score for dimension. This influenced the final score for different dimensions

with varying number of indicators. This issue was addressed by using standardized residuals which

helped to visualize the variation in perceived satisfaction on each indicator under dimensions of

access. This also enabled to observe indicator(s) with large variation in perception within each

dimension, which could be hidden while using summary scores.

From analyses, intra village variation in access to PHC was more prominent than inter village.

Presence of different socioeconomic classes within each village met the assumption about intra

village heterogeneity which to large extent influenced the selection of health care facilities, public or

private. Further the perceived satisfaction with individual factors as well as overall access was greatly

related to the type of facility being visited.

6.2 Sub-objective 2: Existing State of Access in Relation to Health Policy

From the results obtained in section 5.2, existing situation of access to healthcare in study area can be

considered to be in accordance to the current health policy standards of DIY. Physical distances to

healthcare, availability of public health facilities to provide affordable and adequate quality of

service were the main health policy visions in line. Traditional approach in evaluating access to

public services has also given physical accessibility and availability of public facilities vital

importance while measuring service deprivation. These dimensions have also received considerable

attention in the policy aspect of Indonesia at various hierarchies starting from national level, Ministry

of Health, to district level health organizations. All sub districts in DIY contains one or more public

healthcare centres, puskesmas, which was one of the criteria in health policy. Also the ratio of

puskesmas to population as stated in policy was met in present condition. However, there was no

consideration of geographic area of sub districts for allocating health facilities, as there was vast

difference in area and population in sub districts in DIY. Sub districts in city of Yogyakarta were

small with minimum area of 0.65 km2 but highly pupulated and those in rural regencies were large

with maximum area up to 105 km2 with less population in Gugung Kidul. Therefore, geographic

distance to health facilities differs accordingly.

Two sample villages in this study were from urban sub districts in the city of Yogyakarta with areas of

1km2 and 3 km2. The third village was in periphery of the city in Sleman regency with 31 km2 area.

Hence physical accessibility did not seem to be a problematic factor in access to PHC except in few

cases from Sleman sub district. However, situation might be different in larger rural sub districts with

sparse population settlement, although it fulfils the policy standard.

The ratio of puskesmas to population per sub district in the whole province was within the policy

standard which was 1:120,000. But waiting time in puskesmas under availability was found to be one

of the main problems as perceived from household survey also from interviews with medical staffs in

five puskesmas. This refers that the ratio stated in policy norm is large which does not reflect reality,

hence require further attention. Also there was no consideration of ratio of medical staffs to

population in policy standard. Ratio of population to doctor in sample villages was found to be within

the national average ratio (1:7000) in Indonesia, which was again contradicting with the higher

dissatisfaction with long waiting time. It refers to the fact that ratio of health facilities to population

mentioned in policy and the national average ratio of population to doctor were large and unrealistic

when compared to international standards. Although WHO mentions that there are no specific

standards for assessing the sufficiency of medical personnel to address the healthcare need of a given


69

population, average ratio of physician to population was estimated (1:1000) in accordance to achieve

adequate coverage rates for primary healthcare. This ratio was met in only four sub districts out of 78

in the DIY. Although the implementation of existing policies regarding availability of one puskesmas

for every 120,000 look optimistic, however if looked into international health standards, the standard

does not seem to be applicable.

Health policy concerning government health insurances was found to be well implemented primarily

focusing on poor population. By providing free primary healthcare service including medicine cost,

issue of equitable access in obtaining adequate primary healthcare was ensured at macro level.

However, service quality in terms of inter personal treatment was a remarkable problematic issue

found in this study. Although poor people are getting free PHC service, they expressed the feeling of

inequality when it comes to personal behaviour of medical staffs in public health facilities. As a

result, majority of people wish to change their health facility to a private clinic if applicable.

If access to PHC was evaluated only considering the factors mentioned in health policy of DIY, then

result might have shown higher level of access than what was found by this study by measuring five

dimensions of access. Hence evaluating access by considering such dimensions have demonstrated

underlying variations that exists in present situation which might not be in policy focus.

6.3 Sub-objective 3: Variations in Results Obtained from Census and Primary

Data

Comparative analyses between aggregated census data and primary household data showed

significant variation in socioeconomic composition of villages. Census presented very small

percentage of households with the letter of poor and health card in sample villages as compared with

the findings from household survey (refer section 5.1). From this observation it can be said that in

comparison, aggregated census data overestimated the socioeconomic condition of sampled

households, over-shading the heterogeneity within them. Such information on census might not reflect

the real situation when it is being used for studies, where socioeconomic characteristics of people play

a vital role. Also such aggregated data might not provide accurate information while making

purposive selection of study area as in this study.

Loss of variation while aggregating can be one of the reasons behind such variation. Also, it could be

that census and primary data were collected in different ways for different purposes. Sample size used

might be small and by chance, may not be a good representative of the villages. Other reasons could

be that the sample villages in this study might not be very representative and could be an extreme

case, although unlikely. Time gap in collection of information could be another factor in such

variation, as the census data was from 2005 and primary data was collected in 2009.

Variation in the geographic distance from villages to nearest health facilities stated in census and that

obtained from survey was expected. As census simply considered distance from village centre to the

nearest facility without consideration to the mode of transportation, assuming people will visit the

nearest one. But in reality, this distance varies as people have their own preference in choosing a

health facility. Decision of not choosing the nearest facility is mostly related to their dissatisfaction

with certain factors in that facility. This could be an indication of some drawbacks in such health

facilities seeking for improvements.

Higher variations within villages than its sub districts refer that intra village variation can get lost in

aggregation process. Hence studying variation in access to healthcare at lower level, than

administratively defined village boundaries can provide better understanding of problems in various

dimensions of access to healthcare for improvement.

70

7. Conclusions and Recommendations

This chapter gives an overview on the main building blocks of this study starting from the

theoretical conceptualisation to the methodological operationalisation in achieving answers to all

research questions formulated in this study. It is composed in two main sections. First section

presents the conclusive remark on the main findings of this study. Second section proposes

recommendations on possible aspects or directions for further research development based on the

findings and limitations of this study.

7.1 Conclusions

This study was primarily concerned with the evaluation of access to primary healthcare considering

various physical, financial and social factors using province of Yogyakarta as a case study. Unlike

common approach of evaluating access by focusing merely on physical accessibility and availability,

this study conceptualised and operationalised access as consisting of five dimensions developed as

follows: accessibility; availability; affordability; acceptability and adequacy. To achieve the main

objective of this study, three supporting objectives were formulated along with number of research

questions. Coming subsection draws conclusion over the findings of this study in order to address the

research questions to achieve the main objective.

7.1.1 Main Findings from Sub-objective 1

- Analyses to address research questions of first sub objective showed that intra village

variation in access to PHC was higher than inter village due to socioeconomic heterogeneity within

each sample villages. Although variation was not large, Tegalpanggung had slightly lower satisfaction

with access when compared to other two villages, as it contained comparatively large proportion of

poor households. Study of higher spatial resolution than at administratively defined area boundaries,

for instance village in this study, can give better insight of the situation on access to PHC. Also if

areas with high socioeconomic homogeneity can be identified for evaluation, then results of analyses

might give clear variation between such areas.

- By quantifying and measuring relevant indicators under each dimension of access, it was

realized that perceived importance of different factors or dimensions in access was related to the

socioeconomic characteristics of individual to a large extent. Healthcare cost seemed more important

than fast service and better personal treatment basically among lower socioeconomic class. As despite

of dissatisfaction with these issues people continued to visit public health facilities for free or cheap

service cost. In contrary, despite of comparably high service and medicine cost, mostly high

socioeconomic class preferred to visit private facilities for faster and better service. Hence

availability and adequacy, as defined in this study, seemed to be more important dimension than

affordability for well off people. Noticeable variation was observed in overall satisfaction level with

access to PHC between different socioeconomic classes. Majority of unsatisfied respondents were

from lower socioeconomic class and satisfied from higher class. However, the satisfaction level was

related to the type of facilities being visited rather that the socioeconomic class. Therefore while

evaluating variation in access to PHC, type of facility should be considered, which can provide better

indication of problematic factors related with certain type of health facility for further improvement.

- Satisfaction level with access to PHC was found to be influenced by satisfaction with

individual factors under dimensions. As in most of cases, despite of higher satisfaction with majority

of factors under five dimensions of access, the overall satisfaction level was lower (34%) than the


71

dissatisfaction level (48%) in total respondent. In general, factors related to accessibility,

affordability and acceptability were found to be less problematic in study area, as compared to the

long waiting time under availability and dissatisfaction with inter personal treatment by medical

staffs under adequacy. Only these two factors were found to be associated with the overall satisfaction

level of people visiting public health facilities. Feeling of inequality in personal treatment by medical

staffs under adequacy was expressed by lower socioeconomic class visiting public healthcare, which

was related to their dissatisfaction. Other factors like cost, distance and trust on medical ability or

quality of service showed importance while evaluating access to private health facilities. Short

waiting time in facility, good inter personal treatment and high trust on medical ability seemed to be

important reasons for people to visit private clinics. Hence, their satisfaction level with access had

positive relation with their perception over these issues. This refers to the fact that all these issues are

relevant and should be considered while evaluating access to healthcare in general.


- Second sub objective was to study the health planning system and policy standards in

context of Yogyakarta. Aim was to compare existing situation of access to healthcare at micro level

with the policy standards at macro level. Referring to the health policy of the province of

Yogyakarta, the state of access to PHC was found to be in accordance to the existing policy norms.

Access related policy for healthcare was focused on factors like availability of public health

facilities; physical accessibility; affordability and adequate quality of services. As a result of efficient

policy implementation in these aspects, the existing situation of access was found to be in better state.

Accessibility and affordability ranked higher summary scores, also in perceived level of satisfaction

when evaluated at micro level. Limitation in policy intervention was found in case of perceived

problems with availability of medical personnel and in addressing inequality experienced by people

from different socioeconomic strata in terms of inter-personal treatment by medical staffs.

Availability was perceived as the ratio of population to doctors in this study. Although this ratio was

in line with the average population doctor ratio of Indonesia, it looked very large and inapplicable

when compared with international health standards like WHO. Hence, further attention is required at

policy making level in related problematic issues with consideration of international health standards

if applicable. If issues related to personal treatment along with long waiting time in public health

facilities can be addressed from policy level, then existing access to public healthcare can score

remarkably higher satisfaction level at micro level.


Third sub objective of this study was to evaluate the consistency in results obtained from different

scale of analysis. To address this sub objective, comparative analyses was done between common

socioeconomic and accessibility variables, using aggregated census data from 2005 and

disaggregated primary data obtained from household survey. Large difference in socioeconomic and

accessibility variables was observed. Comparative analyses between socioeconomic attributes

demonstrated that aggregated census data often overestimated the socioeconomic characteristics of

people at village level. Also variation found in socioeconomic and accessibility attributes within

villages were reduced or lost when checked at higher administrative unit than village. These results

indicated that aggregation at larger scale, i.e. village, can average out the variability in

socioeconomic and accessibility variables that exists at small scales, i.e. socioeconomic clusters

within villages. In general, the findings indicate that large scale study might hide the variability in

access to PHC at small scales. If homogeneous socioeconomic clusters can be used as the scale of

analysis rather than census data at village level, then the evaluation of access to healthcare can

72

provide more realistic results. Care should be given to the fact that scale and configuration of spatial

units may affect the outcome of analyses. Therefore care should be taken while drawing conclusion or

in decision making when aggregated data are used.

7.2 Recommendations

Following the findings and methodological limitations faced in this study, number of possible areas

can be recommended for future research development that can further enhance the findings of this

study.

- As this study was carried out in the areas located in city core and immediate periphery of the

city, large variation could not be seen in the state of access to PHC between sample villages. Rural

area with sparsely located settlement can give more space to compare the variation in access between

urban and rural regions within one province under similar health policies. Detail spatial analysis, for

example using road network or gravity based models, can be another practical approach to evaluate

accessibility measures under access.

- Another possible field can be to evaluate variation in access at different scale of analysis

using same source of data. As in this study, comparison in aggregated and disaggregated data was

done using secondary census data and primary household data respectively. Evaluating the

consistency of statistical results obtained from aggregated and disaggregated data will be more

precise and accurate if same source of data can be used.

- Following the result of this study that showed variation in access between different

socioeconomic classes, new zone design technique can be proposed for further research development.

This technique can be used to create new zone boundaries by considering maximum internal

socioeconomic homogeneity. Rather than limiting analyses within administratively defined

boundaries, i.e. village, this technique can show more effective and realistic variations in access to

healthcare across different socioeconomic classes. It also allows to observe the effect of modifiable

areal unit problem.

As a final remark of this study, even the most powerful diagnostic tests, medicines and existence of

supreme quality healthcare service cannot improve peoples‟ health status, if they do not reach to

needy people. To ensure equal access in all dimensions to health services by people, regardless of

their socioeconomic strata, is a major challenge in itself. Hence additional efforts should be made

from policy level to enable all population to gain equitable access to primary healthcare. Also health

policies should be formulated in a way to meet the minimum threshold lines for relevant factors.

Consideration to internationally accepted healthcare standards could be more rational in achieving

adequate access to primary health care.

It is expected that the findings of this study, have been useful to define and measure access to primary

healthcare from broader perspective, highlighting the existing underlying problematic factors for

further improvement and enhancement of access to healthcare in the region.


73

References

Amer, S. (2007). Towards spatial justice in urban health services planning : A spatial - analytic GIS -

based approach using Dar es Salaam, Tanzania as a case study.299 pp, Utrecht University.

Andersen, R. M., A. McCutcheon, et al. (1983). "Exploring Dimensions of Access to Medical Care."

Bagheri, N., G. Benwell, L., et al. (2005). "Measuring spatial accessibility to primary health care."

Black, M., S. Ebener, et al. (2004). Using GIS to Measure Physical Accessibility to Health Care.

International Health Conference World Health Organization.

Campbell, S. M., M. O. Roland, et al. (2000). "Defining quality of care." Social Science & Medicine

51(11): 1611-1625.

Carstairs, V. (1981). "Small area analysis and health research." Community Medicine 3: 131-139.

Cockings, S. and D. Martin (2005). "Zone design for environment and health studies using pre-

aggregated data." Social Science & Medicine 60(12): 2729-2742.

Das, D. (2008). "Urban Quality of Life: A Case Study of Guwahati." Social Indicators Research

88(2): 297-310.

Dash, L. N. (2000). World Bank and economic development of India. New Delhi, APH Publ. Corp.

ESCAP (2009). Statistical Yearbook for Asia and the Pacific 2008, United Nations.

Field, A. (2005). Discovering statistics using SPSS. London etc., Sage.

Foo, T. S. (2000). "Subjective assessment of urban quality of life in Singapore (1997-1998)." Habitat

International 24(1): 31-49.

Fosu, G. B. (1989). "Access to Health Care in Urban Areas of Developing Societies." Journal of

Health and Social Behavior 30(4): 398-411.

Fosu, G. B. (1989). "Access to Health Care in Urban Areas of Developing Societies."

Frankenberg, E., J. C. Caldwell, et al. (2004). "The effects of access to health care on infant

mortality in Indonesia."

Guagliardo, M. F. (2004). "Spatial accessibility of primary care: concepts, methods and challenges."

International Journal of Health Geographics 3.

Gulliford, M., J. Figueroa-Munoz, et al. (2002). "What does „access to health care‟ mean?" Journal of

Health Services Research and Policy 7(3): 186–188.

Gulzar, L. (1999). "Access to Health Care." Journal of Nursing Scholarship 31(1): 13-19.

Haynes, R., K. Daras, et al. (2007). "Modifiable neighbourhood units, zone design and residents'

perceptions." Health & Place 13(4): 812-825.

Higgs, G. (2005). "A Literature Review of the Use of GIS-Based Measures of Access to Health Care

Services " Health Services and Outcomes Research Methodology 5(2): 119-139.

Hyera, M. B. (2003). Assessment of accessibility to elementary education in informal settlements :

the case of Cebu city, the Philippines. Enschede, ITC: 39.

Ibrahim, M. F. and S. W. Chung (2003). "Quality of Life of Residents Living near Industrial Estates

in Singapore." Social Indicators Research 61(2): 203-225.

74

Lahelma, E., P. Martikainen, et al. (2004). "Pathways between socioeconomic determinants of

health." J Epidemiol Community Health 58: 327–332.

Leisinger, K., M. (2008). Annual report 2008 Creating access – for better health, Novartis

Foundation for Sustainable Development.

Martin, D., A. Nolan, et al. (2001). "The application of zone-design methodology in the 2001 UK

Census." Environment and Planning A 33(11): 1949-1962.

Martinez, J. A. (2005). Monitoring intra - urban inequalities with GIS - based indicators : with a case

study in Rosario, Argentina, Utrecht University ITC.

Maxwell, R. J. (1992). "Dimensions of quality revisited: from thought to action." 1(3): 171-177.

May, R., K. Rex, et al. (2000). "UN Habitat Indicators Database: evaluation as a source of the status

of urban development problems and programs." Cities 17(3): 237-244.

McGrail, M. R. and J. S. Humphreys (2008). "Measuring spatial accessibility to primary care in rural

areas: Improving the effectiveness of the two-step floating catchment area method." Applied

Geography In Press, Corrected Proof.

Millman, M. L. (1993). Access to health care in America, Institute of Medicine.

Moeloek, F. A. (1999). Health Development Plans Towards Healthy Indonesia 2010, Ministry of

Health, Republic of Indonesia.

Moseley, M. J. (1979). Accessibility: the rural challenge, London, Methuen and Co.

Obrist, B., N. Iteba, et al. (2007). "Access to Health Care in Contexts of Livelihood Insecurity: A

Framework for Analysis and Action." PLoS Med 4(10): e308.

Oliver, A. and E. Mossialos (2004). "Equity of access to health care: outlining the foundations for

action."

Omer, I. (2006). "Evaluating accessibility using house-level data: A spatial equity perspective."

Computers, Environment and Urban Systems 30(3): 254-274.

Parker, A. W., J. M. Walsh, et al. (1976). "A Normative Approach to the Definition of Primary

Health Care." The Milbank Memorial Fund Quarterly. Health and Society 54(4): 415-438.

Parnell, S. and M. Poyser (2001). The value of indicators as a tool for local government,

Democratising Local Government: The South African Experiment. University of Cape Town Press,

Cape Town. pp. 251-260.

Penchansky, R. and J. W. Thomas (1981). "The Concept of Access. Definiton and Relationship to

Consumer Satisfaction, Medical Care." 19(2): 127-140.

Rogerson, P. A. (2006). Statistical methods for geography : a student's guide. London etc., Sage.

Roghmann, K. J., A. Hengst, et al. (1979). Satisfaction with Medical Care: Its Measurement and

relation to Utilization, Medical Care.

Schuurman, N., N. Bell, et al. (2007). "Deprivation Indices, Population Health and Geography: An

Evaluation of the Spatial Effectiveness of Indices at Multiple Scales." Journal of Urban Health 84(4):

591-603.

Smith, D. M. (1994). Geography and Social Justice, Colchester, Blackwell.

Stafford, M., O. Duke-Williams, et al. (2008). "Small area inequalities in health: Are we

underestimating them?" Social Science & Medicine 67(6): 891-899.


75

Talen, E. and L. Anselin (1998). "Assessing spatial equity: an evaluation of measures of accessibility

to public playgrounds." Environment and Planning A 30: 595-613.

Tsou, K. W., Y. T. Hung, et al. (2005). "An accessibility-based integrated measure of relative spatial

equity in urban public facilities." Cities 22(6): 424-435.

Turksever, A. N. and G. Atalik (2001). "Possibilities and Limitations for the Measurement of the

Quality of Life in Urban Areas." Social Indicators Research 53(2): 163-187.

Unger, A. and L. W. Riley (2007). "Slum Health: From Understanding to Action." PLoS Med 4(10):

e295.

Veenhoven, R. (2002). "Why Social Policy Needs Subjective Indicators." Social Indicators Research

58(1): 33-46.

Wagstaff, A. (2002). "Poverty and health sector inequalities." Bulletin of the World Health

Organization 80: 97-105.

WHO (2009). "Indonesia Country Health Profile - Development of the health system." Retrieved 22-

12-2009, 2009, from www.searo.who.int/EN/Section313/Section1520_6822.htm.

Wong, C. (2006). Indicators for urban and regional planning : the interplay of policy and methods :

also e-book. London etc., Routledge.

Yamauchi, F., S. Chowdhury, et al. (2007). "Spatial coordination in public good allocation:

Nonparametric evidence from decentralized Indonesia " International Food Policy Research Institute

(IFPRI)

http://www.searo.who.int/EN/Section313/Section1520_6822.htm

76

Appendix A: Empirical references to indicators used in this study

Indicators

Pen

chan

sky a

nd

Th

om

as

(1981)

An

der

sen

, M

cCu

tch

eon

et a

l. (

1983)

Fosu

, G

.B.

(19

89)

May,

Rex

et

al.

(2

000)

Wagst

aff

(2002)

Gu

agli

ard

o (

2004)

Fra

nk

enb

erg,

Cald

wel

l

et a

l.

(2004)

Oli

ver

an

d M

oss

ialo

s

(2004)

Un

ger

an

d R

iley

(2007)

Ob

rist

, It

eba e

t al.

(2007)

Am

er (

2007)

Lei

sin

ger

(2008)

Gu

lzar

(1999)

Socio-

economic

Attributes

Household

characteristics

Education level √ √ √ √ √ √ √ √ √ √ √ √

Age dependency √ √ √ √ √

Employment

status /

dependency

√ √ √ √ √

Household

income

√ √ √ √ √ √ √ √ √ √ √ √

Housing

condition

House ownership √ √ √ √ √ √

Land ownership √ √ √

Construction

type

√ √ √ √ √

House crowding √ √ √

Physical

infrastructure

Toilet type √ √ √ √ √ √ √

Source of water √ √ √ √ √ √ √

Electricity √ √ √ √

Sewage disposal √ √ √ √ √

Assets

possession

Type of vehicles √ √ √ √

Telephone √

Television √

Refrigerator √

Dimensions of Access to PHC

Availability Type of PHC

facility

√ √ √ √ √ √ √ √ √ √

PHC facility-

population ratio

√ √ √ √

Doctor -

population ratio

√ √ √ √ √ √

Medical drug

store

√ √ √ √ √

Ease to get

appointment

√ √

Waiting time to

get check-up

√ √ √ √ √ √

Subjective

perception

√ √ √ √ √


77

Indicators

Dimensions of Access to PHC

Pen

chan

sky a

nd

Th

om

as

(1981)

An

der

sen

, M

cCu

tch

eon

et a

l. (

1983)

Fosu

, G

.B.

(1989)

May,

Rex

et

al.

(2

000)

Wagst

aff

(2002)

Gu

agli

ard

o (

2004)

Fra

nk

enb

erg,

Cald

wel

l

et a

l.

(2004)

Oli

ver

an

d M

oss

ialo

s

(2004)

Un

ger

an

d R

iley

(2007)

Ob

rist

, It

eba e

t al.

(2007)

Am

er (

2007)

Lei

sin

ger

(2008)

Gu

lzar

(1999)

Accessibility

Travel

distance

√ √ √ √ √ √ √ √ √

Travel time √ √ √ √ √ √ √ √ √ √

Mode of

transport

√ √ √ √ √ √ √ √ √

Subjective

perception

√ √ √ √ √ √

Affordability Possession of

hearth card Or

health

insurance

√ √ √ √ √ √

Insurance

coverage

√ √ √ √ √

Direct and

indirect costs

√ √ √ √ √ √ √ √ √

Subjective

perception

√ √ √ √ √ √ √

Acceptability Religious or

cultural

factors

√ √ √

Adequacy Quality of

medical care

√ √ √ √ √ √ √ √ √

Personal

treatment

√ √ √ √ √ √ √ √

Cleanliness √ √ √ √

Opening hour √ √ √ √

Satisfaction level with access to

health care

√ √ √ √ √

EV

AL

UA

TIO

N O

F A

CC

ES

S T

O P

RIM

AR

Y H

EA

LT

HC

AR

E, C

AS

E S

TU

DY

IN

YO

GY

AK

AR

TA

79

Ap

pe

nd

ix B

: D

es

cri

pti

on

an

d r

ati

on

ale

s o

f in

dic

ato

rs u

se

d i

n t

his

stu

dy

Ind

ica

tors

A

ttri

bu

tes

Des

crip

tio

n

Rati

on

ale

So

cio

-

eco

no

mic

Ho

use

ho

ld

char

acte

rist

ics

Hig

hes

t

educa

tion l

evel

Hig

hes

t le

vel

of

edu

cati

on

att

ain

ed b

y

house

ho

ld m

emb

ers

“Ed

uca

tio

n p

rovi

des

form

al q

ual

ific

atio

ns

that

contr

ibute

to t

he

soci

oec

onom

ic s

tatu

s

thro

ugh

occ

up

atio

n a

nd i

nco

me”

(L

ahel

ma,

Mar

tikai

nen

et

al.

2004).

In t

his

stu

dy i

t

con

trib

ute

s to

ob

serv

e th

e in

fluen

ce o

f ed

uca

tion l

evel

on t

he

atti

tude

of

house

hold

tow

ard

s h

ealt

hca

re a

nd

thei

r per

cepti

on o

n d

iffe

rent

fact

ors

rel

ated

to t

he

acce

ss t

o P

HC

.

Age

dep

enden

cy

Nu

mer

ato

r:

Ho

use

ho

ld m

emb

ers

wit

hin

age

of

16

- 6

0 y

ears

old

Den

om

ina

tor:

To

tal

ho

use

ho

ld m

emb

ers

Peo

ple

bet

wee

n 1

6 –

60

yea

rs (

refe

rrin

g t

o c

ensu

s d

ata

and

off

icia

l re

tire

men

t age

in

Ind

on

esia

) ar

e as

sum

ed t

o b

e ac

tive

and a

ble

to w

ork

for

earn

ing.

House

hold

conta

inin

g

mo

re m

emb

ers

bet

wee

n t

his

age

gro

up c

an h

ave

more

poss

ibil

ity i

n i

mpro

ving t

he

eco

no

mic

sta

tus.

Em

plo

ym

ent

dep

enden

cy

Nu

mer

ato

r:

Fu

ll t

ime

emp

loyee

+ p

art

tim

e em

plo

yee

/ 2

Den

om

ina

tor:

To

tal

ho

use

ho

ld m

emb

ers

Ho

use

ho

ld e

arn

ing i

s der

ived

pri

mar

ily f

rom

pai

d e

mplo

ym

ent.

Ho

use

ho

ld w

ith

lar

ge

num

ber

of

emplo

yed

mem

ber

s ca

n b

e as

sum

ed t

o h

ave

more

ear

nin

g

and

bet

ter

eco

no

mic

sta

tus.

Ran

ge

of

inco

me

Ran

ge

of

ho

use

ho

ld m

on

thly

in

com

e, i

n

Indo

nes

ian

Ru

pia

h

1.

Les

s th

an 7

00

,00

0

2.

70

0,0

00

– 1

,40

0,0

00

3.

1,4

00

,00

0 –

2,8

00

,00

0

4.

2,8

00

,00

0 –

5,6

00

,00

0

5.

Ab

ove

5,6

00

,00

0

Inco

me

det

erm

ines

th

e purc

has

ing p

ow

er o

f house

hold

. It

als

o c

ontr

ibute

s to

res

ourc

es

nee

ded

in

mai

nta

inin

g g

ood h

ealt

h.

Inco

me

bei

ng a

sen

siti

ve m

atte

r, a

ran

ge

is u

sed t

o p

ut

resp

onden

t at

eas

e an

d a

lso t

o a

void

ove

rest

imat

ion

or

un

der

esti

mat

ion t

o s

om

e ex

tent.

(Ref

erri

ng

to

Sta

tist

ics

of

Ind

ones

ia 7

00,0

00 I

nd

ones

ian R

upia

h i

s co

nsi

der

ed a

s th

e

inco

me

po

vert

y li

ne

for

a h

ouse

hold

wit

h 4

-5 m

ember

s.)

Poss

essi

on o

f

„let

ter

of

poor‟

Ask

ing i

f th

e h

ou

seh

old

hav

e th

e le

tter

of

poor

It i

s an

off

icia

l le

tter

iss

ued

by t

he

loca

l gove

rnm

ent

whic

h i

s an

indic

atio

n o

f a

poor

ho

use

ho

ld i

n c

on

tex

t o

f In

dones

ia.

Ho

usi

ng

con

dit

ion

House

and L

and

ow

ner

ship

Sta

tus

of

ho

use

an

d l

and

ow

ner

ship

;

ow

ned

, re

nte

d o

r sq

uat

ted

Ow

ner

ship

of

ho

use

and l

and e

nsu

res

less

fin

anci

al b

urd

en o

n h

ouse

hold

in t

erm

s of

mo

nth

ly r

ent.

Const

ruct

ion

type

Co

nst

ruct

ion

typ

e o

f th

e h

ou

se, if

it

is

per

man

ent,

sem

i p

erm

anen

t o

r te

mp

ora

ry

Ph

ysi

cal

con

dit

ion

an

d t

ype

of

const

ruct

ion c

an e

xpre

ss t

he

livi

ng s

tandar

d o

f house

hold

mem

ber

s.

80

stru

ctu

re

Def

init

ion

of

con

stru

ctio

n t

ypes

fro

m c

ensu

s ques

tionnai

re:

-Per

man

ent

ho

use

is

the

one

wit

h b

rick

, co

ncr

ete

or

wooden

wal

ls,

roof

mad

e out

of

alu

min

ium

sh

eets

or

wooden

sla

tes

and f

loors

wit

h c

oncr

ete

or

cera

mic

til

es.

-Sem

i p

erm

anen

t h

ou

se a

re m

ade

wit

h h

alf

concr

ete,

wood o

r bam

boo w

alls

wit

h r

oofs

mad

e o

f ti

les

or

alu

min

ium

shee

ts o

r w

ood o

r as

bes

tos.

-Tem

po

rary

or

sim

ple

house

are

those

mad

e out

of

mud,

wood a

nd l

eave

s.

C

row

din

g

Rat

io o

f to

tal

nu

mb

er o

f b

edro

om

s to

th

e

tota

l h

ou

seh

old

mem

ber

s

Nu

mb

er o

f p

eop

le p

er b

edro

om

is

calc

ula

ted t

o s

ee t

he

infl

uen

ce o

f bed

shar

ing a

s it

is

con

sid

ered

as

a se

nsi

tive

cro

wdin

g i

ndic

ator

also

for

hea

lth s

tudie

s.

Ph

ysi

cal

infr

astr

uct

ure

Toil

et t

ype

Typ

e o

f to

ilet

use

d, if

it

is p

riva

te, p

ub

lic

(sh

ared

) o

r n

o t

oil

et (

for

exam

ple

, in

rive

r)

Bas

ic i

nfr

astr

uct

ure

s li

ke

wat

er s

upply

, sa

nit

atio

n a

nd e

lect

rici

ty a

re c

om

monly

use

d

ind

icat

or

for

soci

oec

onom

ic s

trat

ific

atio

n.

Bei

ng b

asic

nee

ds

for

dai

ly a

ctiv

itie

s, t

hes

e in

fras

truct

ure

s re

flec

t th

e li

ving s

tandar

d a

lso

the

hygie

ne

con

dit

ion

of

house

hold

mem

ber

s. A

s so

urc

e of

dri

nkin

g w

ater

supply

and t

ype

of

san

itat

ion

is

dir

ectl

y r

elat

ed w

ith h

ealt

h s

tatu

s.

Sourc

e of

wat

er

If h

ou

seh

old

hav

e p

ub

lic

pip

ed w

ater

supp

ly (

PA

M),

wel

l, ra

inw

ater

or

fro

m

rive

r as

th

e so

urc

e o

f w

ater

Ele

ctri

city

If

ho

use

ho

ld h

ave

pu

bli

c el

ectr

ic s

up

ply

(PL

N),

so

me

oth

er t

yp

e o

r n

o e

lect

rici

ty

Ass

ets

po

sses

sio

n

Num

ber

of

vehic

le

Num

ber

of

bic

ycl

es, m

oto

r b

ikes

or

cars

A

sset

po

sses

sio

n i

s an

oth

er s

oci

oec

onom

ic i

ndic

ator

whic

h i

ndic

ates

the

stat

us

and a

bil

ity

of

ho

use

ho

ld t

o m

ain

tain

cer

tain

lev

el o

f li

ving s

tandar

d.

As

inco

me

bei

ng a

sen

siti

ve a

nd s

om

etim

es c

onfi

den

tial

mat

ter,

eva

luat

ion o

f su

ch a

sset

s

can

be

a u

sefu

l w

ay f

or

soci

oec

onom

ic s

trat

ific

atio

n.

Num

ber

of

phone;

TV

;

refr

iger

ator

Num

ber

of

such

ho

use

ho

ld a

sset

s o

wn

ed

by t

he

ho

use

ho

ld

Acc

ess

to P

HC

Ava

ilab

ilit

y

Deg

ree

of

fit

bet

wee

n

exis

tin

g P

HC

serv

ices

and

nee

ds

of

peo

ple

Type

of

PH

C

faci

lity

Typ

e o

f P

HC

fac

ilit

y v

isit

ed, if

it

is

pu

bli

c(go

vern

men

t) o

r p

riva

te

org

aniz

atio

n

Typ

e o

f P

HC

fac

ilit

y b

eing v

isit

ed c

an b

e use

ful

to e

valu

ate

resp

onden

t‟s

per

cepti

on

tow

ard

s va

rio

us

fact

ors

rel

ated

to h

ealt

hca

re s

ervi

ce.

It c

an a

lso e

xpre

ss t

he

soci

oec

onom

ic

stat

us

of

resp

on

den

t. A

s in

conte

xt

of

Indones

ia,

gove

rnm

ent

hea

lth c

entr

es p

rovi

de

free

PH

C s

ervi

ce t

o p

oo

r p

eople

.

PH

C f

acil

ity-

popula

tion r

atio

Num

ber

of

PH

C f

acil

itie

s d

ivid

ed b

y

tota

l p

op

ula

tio

n i

n e

ach

su

b-d

istr

ict

Th

e h

ealt

h p

oli

cy i

n I

nd

ones

ia h

ave

spec

ifie

d a

min

imum

num

ber

of

hea

lthca

re f

acil

itie

s

per

ad

min

istr

ativ

e u

nit

s

Th

e ra

tio

of

faci

lity

or

doct

ors

to p

opula

tion s

how

s th

e dem

and

-supply

rat

io o

f P

HC

serv

ice.

It

also

en

able

s to

com

par

e th

e ex

isti

ng s

ituat

ion w

ith t

he

DIY

poli

cy a

nd

Doct

or

-

po

pula

tion r

atio

Tota

l n

um

ber

of

do

cto

rs d

ivid

ed b

y t

ota

l

popu

lati

on

in

eac

h s

ub

-dis

tric

t

EV

AL

UA

TIO

N O

F A

CC

ES

S T

O P

RIM

AR

Y H

EA

LT

HC

AR

E, C

AS

E S

TU

DY

IN

YO

GY

AK

AR

TA

81

inte

rnat

ion

al h

ealt

h s

tandar

ds.

Med

ical

dru

g

store

Ava

ilab

ilit

y o

f m

edic

al s

tore

in

PH

C

faci

lity

vis

ited

Ava

ilab

ilit

y o

f d

rug s

tore

for

pre

scri

bed

med

icin

e is

an i

mport

ant

fact

or

consi

der

ed i

n

man

y e

mp

iric

al s

tud

ies

(Wag

staf

f 2002;

Am

er 2

007;

Obri

st,

Iteb

a et

al.

2007)

whil

e

eval

uat

ing a

cces

s to

hea

lthca

re.

It i

s as

sum

ed t

hat

peo

ple

s‟ o

pin

ion o

n o

vera

ll a

vail

abil

ity

of

hea

lth

care

is

infl

uen

ced b

y h

ow

eas

y o

r dif

ficu

lt i

t is

to g

et r

equir

ed m

edic

ine

is.

Eas

e to

get

appoin

tmen

t

Do t

hey

hav

e to

get

an

ap

po

intm

ent

pri

or

visi

tin

g t

he

faci

lity

or

just

wal

k i

n?

If

yes

, h

ow

lo

ng d

oes

it

tak

e to

get

it?

Qu

ery a

bo

ut

app

oin

tmen

t sy

stem

aid

to u

nder

stan

d t

he

pro

cess

by w

hic

h p

eople

can

ente

r

dif

fere

nt

hea

lth

care

fac

ilit

y t

o o

bta

in t

he

serv

ice.

Als

o t

o k

now

how

long t

hey

hav

e to

wai

t

bef

ore

act

ual

ly e

nte

rin

g t

he

faci

lity

, if

appoin

tmen

t is

nee

ded

.

Wai

ting t

ime

to

get

chec

k-u

p

Wai

tin

g t

ime

afte

r re

gis

trat

ion

in

th

e

faci

lity

til

l th

e v

isit

to

do

cto

r fo

r ch

eck

-

up

Wai

tin

g t

ime

in h

ealt

hca

re f

acil

ity i

s an

im

port

ant

and c

om

monly

use

d i

ndic

ator

in

hea

lth

care

rel

ated

stu

die

s. D

espit

e of

physi

cal

avai

labil

ity o

f se

rvic

e ce

ntr

e, p

atie

nts

mig

ht

hav

e to

wai

t lo

ng t

o o

bta

in m

edic

al s

ervi

ce d

ue

to i

nad

equat

e per

sonnel

, eq

uip

men

t or

som

e o

ther

rea

son

. H

ence

, th

is i

ndic

ator

can g

ive

an i

ndic

atio

n o

f re

lati

on b

etw

een v

olu

me

of

hea

lth

care

ser

vice

dem

and a

nd s

upply

.

Su

bje

ctiv

e p

erce

pti

on

on w

aiti

ng t

ime

is r

elev

ant

in t

his

stu

dy a

s it

was

ass

um

ed t

hat

per

cep

tio

n m

igh

t d

iffe

r dep

endin

g o

n s

oci

oec

onom

ic c

har

acte

rist

ics

of

peo

ple

and a

lso

dep

end

ing o

n t

he

typ

e o

f hea

lthca

re f

acil

ity v

isit

ed.

Subje

ctiv

e

per

cepti

on

Opin

ion

ab

ou

t th

e w

aiti

ng t

ime

in

faci

lity

bef

ore

get

tin

g c

hec

ku

p i

n f

ive

Lik

ert

scal

e; V

ery s

ho

rt. S

ho

rt, N

orm

al,

Lon

g a

nd

Ver

y l

on

g

Acc

essi

bil

ity

Deg

ree

of

fit

bet

wee

n

geo

gra

phic

al

loca

tio

n

of

PH

C s

ervi

ce

an

d th

e

loca

tio

n o

f

peo

ple

Is t

he

visi

ted

faci

lity

the

nea

rest

one

Do t

he

ho

use

ho

ld v

isit

th

e n

eare

st P

HC

faci

lity

It w

as a

ssu

med

th

at p

eople

are

rat

ional

and v

isit

s th

e nea

rest

PH

C f

acil

ity i

n i

dea

l ca

se.

Th

is i

nd

icat

or

hel

ps

to e

nsu

re t

he

assu

mpti

on.

Tra

vel

dis

tance

P

hysi

cal

dis

tan

ce t

o t

he

PH

C f

acil

ity

visi

ted

Tra

vel

dis

tan

ce a

nd

tim

e ar

e th

e co

mm

only

use

d i

ndic

ators

to e

valu

ate

trav

el i

mped

ance

in

any a

cces

sib

ilit

y r

elat

ed s

tudie

s. I

t ca

n b

e use

d t

o a

nal

yze

how

far

or

how

long p

eople

are

wil

lin

g t

o t

rave

l to

get

PH

C s

ervi

ce.

Mo

de

of

tran

spo

rt i

s an

oth

er f

acto

r to

be

consi

der

ed i

n t

his

stu

dy a

s it

is

dir

ectl

y r

elat

ed t

o

ph

ysi

cal

acce

ssib

ilit

y. F

or

inst

ance

, non

-moto

rize

, m

oto

rize

d o

r publi

c tr

ansp

ort

atio

n.

Tra

vel

tim

e T

ota

l tr

avel

lin

g t

ime

to r

each

th

e P

HC

faci

lity

Mode

of

tran

sport

Mod

e o

f tr

ansp

ort

use

d t

o v

isit

PH

C

faci

lity

Subje

ctiv

e

per

cepti

on

Opin

ion

ab

ou

t th

e tr

avel

dis

tan

ce a

nd

trav

el t

ime

to r

each

th

e P

HC

fac

ilit

y

Su

bje

ctiv

e p

erce

pti

on

on t

rave

l im

ped

ance

can

hel

p i

n u

nder

stan

din

g b

ehav

ioura

l

dif

fere

nce

in

var

iou

s d

emogra

phic

or

soci

oec

onom

ic c

lass

es o

f peo

ple

.

Aff

ord

abil

ity

Deg

ree

of

fit

Poss

essi

on o

f

hea

rth c

ard

If t

he

ho

use

ho

ld h

ave

the

hea

lth

car

d,

„Kat

u S

ehat

‟

„Kat

u S

ehat

‟ is

a t

yp

e of

gove

rnm

ent

hea

lth i

nsu

rance

pro

vidin

g f

ree

PH

C s

ervi

ce w

hic

h

has

a d

irec

t re

lati

on

wit

h t

he

dim

ensi

on a

fford

abil

ity i

n t

his

stu

dy.

Insu

rance

If

th

e h

ealt

h c

ard

co

vers

med

ical

an

d

Ap

art

fro

m d

irec

t se

rvic

e co

st l

ike

doct

or‟

s fe

e, t

her

e ar

e m

any i

ndir

ect

expen

ses

in

82 bet

wee

n P

HC

serv

ice

cost

an

d

peo

ple

s’

abil

ity

or

wil

lin

gn

ess

to

pa

y

cove

rage

lab

ora

tory

ex

pen

ses

hea

lth

care

wh

ich

sh

ou

ld n

ot

be

ignore

d w

hil

e ev

aluat

ing o

ver

all

acce

ss t

o P

HC

.

Un

der

stan

din

g o

f in

sura

nce

cove

rage

(hea

lth c

ard i

n t

his

cas

e) o

f th

ese

indir

ect

expen

ses

is

rele

van

t, a

s it

can

co

mp

lete

ly i

nfl

uen

ce p

eople

s‟ o

pin

ion o

n h

ealt

hca

re c

ost

.

Dir

ect

and

indir

ect

cost

s

Co

st o

f re

gis

trat

ion

, d

oct

or‟

s fe

e,

med

icin

e, l

abo

rato

ry a

nd

tra

vel.

Subje

ctiv

e

per

cepti

on

Opin

ion

ab

ou

t th

e o

vera

ll c

ost

fo

r P

HC

Acc

epta

bil

ity

Deg

ree

of

fit

bet

wee

n t

he

cha

ract

eris

tics

of

the

pro

vid

er

an

d t

ho

se o

f

peo

ple

Rel

igio

us

or

cult

ura

l fa

ctors

Do r

esp

on

den

t h

ave

any c

ult

ura

l o

r

reli

gio

us

pre

fere

nce

to

vis

it c

erta

in P

HC

faci

lity

?

Peo

ple

mig

ht

hav

e ce

rtai

n r

elig

ious

or

cult

ura

l pre

fere

nce

in v

isit

ing a

par

ticu

lar

hea

lthca

re

faci

lity

, ei

ther

bec

ause

of

som

e so

cial

beh

avio

ur

or

per

sonal

wis

h.

For

inst

ance

, peo

ple

no

rmal

ly w

ish

to

vis

it a

fac

ilit

y w

her

e th

ey c

an c

om

munic

ate

in l

oca

l la

nguag

e. S

imil

arly

,

gen

der

pre

fere

nce

is

anoth

er f

acto

r in

acc

epta

bil

ity,

as i

n s

om

e ca

se f

emal

e pat

ients

wis

h

to b

e ch

eck

ed b

y f

emal

e doct

ors

and m

ale

pat

ients

by m

ale

doct

ors

.

Gen

der

pre

fere

nce

Is t

her

e an

y g

end

er p

refe

ren

ce o

f

doct

ors

?

Ad

equ

acy

Deg

ree

of

fit

bet

wee

n

peo

ple

s’

exp

ecta

tion

an

d s

ervi

ce

pro

vid

ed

Tru

st o

n m

edic

al

qual

ity

Opin

ion

ab

ou

t th

e m

edic

al a

bil

ity a

nd

serv

ice

qu

alit

y o

f th

e vi

site

d P

HC

faci

lity

?

Des

pit

e o

f sh

ort

ph

ysi

cal

dis

tance

to h

ealt

hca

re f

acil

ity o

r bei

ng p

rovi

ded

wit

h f

ree

hea

lth

care

ser

vice

, p

eople

mig

ht

not

be

hap

py w

ith t

he

serv

ice

due

to s

ever

al o

ther

rea

sons.

Fo

r ex

amp

le p

oo

r se

rvic

e qual

ity c

om

par

ed t

o o

ther

hea

lthca

re f

acil

itie

s, u

nfr

iendly

per

son

al t

reat

men

t b

y m

edic

al p

erso

nnel

, unple

asan

t or

dir

ty p

hysi

cal

envi

ronm

ent

etc.

Op

enin

g h

ou

r o

f h

ealt

hca

re s

ervi

ce s

hould

als

o b

e co

nsi

der

ed a

s peo

ple

mig

ht

hav

e

pro

ble

m v

isit

ing t

he

faci

lity

due

to o

verl

appin

g i

n o

pen

ing t

ime

of

the

faci

lity

and t

hei

r

wo

rkin

g (

emp

loym

ent)

sch

edule

.

Su

ch i

nd

icat

ors

are

rel

evan

t in

this

stu

dy a

s th

ey i

nfl

uen

ce t

he

ove

rall

sat

isfa

ctio

n l

evel

of

peo

ple

to

war

ds

acce

ss t

o P

HC

.

Per

sonal

trea

tmen

t

Opin

ion

ab

ou

t th

e p

erso

nal

tre

atm

ent

by

faci

lity

per

son

nel

in

5 L

iker

t sc

ale

(Ver

y

good

, go

od

, n

orm

al, b

ad a

nd

ver

y b

ad)

Cle

anli

nes

s O

pin

ion

ab

ou

t th

e cl

ean

lin

ess

or

physi

cal

app

eara

nce

of

the

visi

ted

fac

ilit

y

Open

ing h

our

Open

ing h

ou

r o

f th

e P

HC

fac

ilit

y a

nd

if

it s

uit

s th

eir

wo

rkin

g h

ou

r

Su

bje

ctiv

e

per

cep

tio

n o

n

ove

rall

acc

ess

to P

HC

Sat

isfa

ctio

n l

evel

T

akin

g a

ll t

he

fact

ors

an

d d

imen

sio

ns

of

acce

ss i

nto

co

nsi

der

atio

n, o

pin

ion

on

th

e

ove

rall

sat

isfa

ctio

n l

evel

to

war

ds

acce

ss

to P

HC

fac

ilit

y

Rat

ion

al o

pin

ion

on

sat

isfa

ctio

n l

evel

tow

ards

acce

ss t

o P

HC

can

ref

lect

the

actu

al

situ

atio

n, in

ter

ms

of

peo

ple

s‟ p

erce

pti

on,

of

acce

ss t

o P

HC

tak

ing e

ach d

imen

sion i

n

acco

un

t.


83

Appendix C: Content of household survey questionnaire

Sub-district: ………………………………..….

Interviewer name: …………………………….

Village: …………………………………………

Date: ……………………………………..…..…

House no. : ……………………………..…….. Duration: ……………………….…………..….

This survey is purely for study purposes to understand the existing condition in accessing health care facilities in

the city of Yogyakarta, Indonesia. The individual perception of citizens is of immense value for a successful

completion of this study. Your responses to this questionnaire will be treated with strict confidentiality. Hence,

your honest comments and cooperation will be highly appreciated.

Socio-economic Information

[* Read: I will start this interview with some questions related to your household information.]

A. General information of respondent

A.1.

Respondent is

Male Female

A.2. Position in family Head of family Yes No

Husband Wife

A.3. Household members

[Tick the box for those

who are currently living in

the house from atleast

past 1 year]

Below age 6 years: ………………………..

Age 7 – 15 years: …….……………………

Age 16 – 45 years: ………………………..

Age 46 – 60 years: ………………………..

Above 60 years: …………………………..

Total number (including respondent): ………..…

A.4 Highest education level in

household

No education Elementary school

Junior high school Senior high school

University education Others

A.5. Employment status of

household members

Household member

(eg. husband / wife / son)

Full time job Part time job

…………………….

…………………….

…………………….

A.6. Average monthly income

of household in

Indonesian rupiah

Less than 700,000

700,000 – 1,400,000

1,400,000 – 2,800,000

2,800,000 – 5,600,000

Above 5,600,000

A.7. Do you have „letter of

poor‟ (Surat Keterangan

Tidak Mampu)?

Yes No

A.8. In which socio-economic

group your household

belongs to?

Sangat kaya (High class) Kaya (Upper middle class)

Menengah (Middle class) Miskin (Lower middle class)

Sangat miskin (Poor)

84

[* Read: Now, I will ask some general questions about your house condition.]

B. Housing Condition

B.9. Status of house Owned Rented Other

B.10

.

Status of land Owned Rented Squatted (informal)

B.11

.

Number of rooms

occupied by household

Bedroom: …………………....… Kitchen: ……………..……….

Toilet / Bathroom: …………..… Others: ………………….……

B.12

.

Construction type Type: Permanent Semi-permanent Temporary

B.13

.

Type of toilet Public Private Open air (no toilet)

Shared Not shared

B.14

.

Wastewater disposal

(sewage waste)

Septic tank Sewer line None Other

B.15

.

Trash disposal (garbage) Garbage collection system (……..……………….how often)

In container Burn Roadside

In river Other ……………….…………..……..

B.16

.

Main source of water

supply in house

PAM piped water to house Buy from vendors

PAM piped water in public tap Ground water (well)

Others………………………………………………………

B.17

.

Which source of

electricity supply you

have in house?

PLN (State electricity company) Local government agency

Private corporation Public self reliance agency

No electricity Others………………………

B.18

.

Which of these assets

does your household

have?

Cycle………………….. Telephone……….………….

Motorbike……………… Television…………....….….

Car…………………….. Refrigerator…………....……

Other mode of transport……………………..

Existing health care facility

[* Read: I will now ask you about health condition in your household and health facility you visit.]

C. General information on access to health care

C.19.

In the last 6 months,

how many times have

your household

members suffered

from diseases

mentioned in the

table at right box?

Disease

Members Age How

many

times

Severe flu (cough, cold)

Fever

Diarrhea & vomiting

Malaria

Measles (Campak)

Dengue fever (Demam Berdarah)

Upper Respirator Tract Infection (ISPA)

Other……………………………………..

C.20. ** Which health

facility do your

household members

visit for primary

health care?

(Tick the boxes) Order of visit

Hospital …………

Health care (Puskesmas) …………

Health sub-care (Puskesmas Pembantu) …………

Integrated health posts (Posyandu) …………


85

If more than one then

please state in order,

which one do you

visit first and then

further, starting from

1 onwards.

Midwife (Bidan) …………

Traditional birth attendance (Paraji) …………

Others …………………………………………

What is the type of facility?

Public (government) Private

Type of treatment.

By professional medical personnel (like doctor)

By traditional practitioner without formal medical education

Name of facility:………… ……………………………………………

C.21. Is the facility visited

by your household,

the nearest one?

Yes No

If no, why didn‟t you go to the nearest one?

More expensive Don‟t trust the ability of doctors

Unfriendly behavior Don‟t like the quality of service

Religious factor Other………………….……………

C.22. How far is the

facility that your

household visits?

Distance: ………………………….

Time to reach: ……….…………… (in normal condition)

…………………… (with traffic jams, if any)

C.23. Which mode of

transport is used to

reach the facility?

Walk Bicycle Motorbike Private car

Public transportation

If public transport, is it easily available? Yes No

How long do you have to wait to get it? ............................... (time)

C.24. What do you think

about the distance to

the facility?

Very near Near Normal Far Very far

C.25. What do you think

about the travel time

to reach the facility?

Very short Short Normal Long Very long

C.26. Is it necessary to get

an appointment?

Yes No

If yes, can it be done by telephone call?

Yes No

Getting appointment is:

Very easy Easy Normal Difficult Very difficult

C.27. How long do you

normally have to wait

to get a checkup,

after reaching the

health facility?

Time: ……………………………………………..

What do you think about the waiting time?

Very short Short Normal Long Very long

Does it have proper waiting area? Yes No

C.28. Does the facility

have a medical shop

providing prescribed

medicines?

Yes No

If no, how far do you have to go to buy the prescribed medicine?

Very near Near Normal Far Very far

C.29. Does your household

have a health card

(Kartu Sehat) or

health insurance?

Yes No

If yes, who provides it? Government Private organization

Other……………………………………

If no, Why?

86

Not eligible Long process to get the card

Have to travel long distance Unaware

Don‟t need it Other………………………..

C.30. How much do you

have to pay to get

health insurance?

………………………………...……Rupiah Free

If not free, what do you think about the cost?

Very inexpensive Inexpensive Normal

Expensive Very expensive

C.31. ** Does the

insurance cover all

health care

expenses?

Yes No Only primary health care

If no, up to what amount does it cover for bigger diseases?

……………………………..........................

Does it cover all medicine expense? Yes No

C.32. In which health

facilities does the

health card or

insurance give free

health check up?

(See C.20. for

reference)

Government (Public) Private

Hospital

Puskesmas

Puskesmas Pembantu

Posyandu

Midwife (Bidan)

Traditional birth attendance

Hospital

Puskesmas

Puskesmas Pembantu

Posyandu

Midwife (Bidan)

Traditional birth attendance

All

Others ………………………………………………….…………

C.33. How much additional

cost does your

household spend in

health care?

………………… Registration cost ………………… Doctor‟s fee

………………… In medicines …….……….…… Travel cost

…………….…… Overall

C.34.

What does your household think about these costs?

Doctor‟s fee Very inexpensive Inexpensive Normal


Medication cost Very inexpensive Inexpensive Normal


Travel cost to get

health care

Very inexpensive Inexpensive Normal


Total cost Very inexpensive Inexpensive Normal


C.35. Does your household

feel welcome in the

facility you visit?

Yes No

If no, how and why

…………………………………………………………………..………

C.36. Is there any cultural or

religious preference in

choosing a particular

facility?

Yes No

If yes, what……………………………………………………………

Very satisfied Satisfied Normal

Unsatisfied Very unsatisfied

C.37. How is the cleanliness

of the facility?

Very Clean Clean Normal Dirty Very dirty


87

C.38.

How is the personal

treatment from all

facility personnel?

Very Good Good Normal Bad Very bad

C.39.

What does your

household think about

medical ability (trust)

of the facility?

Very Good Good Normal Bad Very bad

**C.40

.

If an equal number of

male and female

medical personnel is

available, to whom

will your household

prefer to visit for

check up?

Male household members to male doctor

Female household members to female doctor

Does not matter for male female

Does not matter at all

How satisfied are you with existing situation in this regard?



C.41. Does the opening

hour of the facility

suits your household

time?

Yes No

C.42.

If your household

income is doubled,

will your household

still go to the same

health care facility?

Yes No

If no, then which one (name)

………………………………………………………………….…….…

Why?………………………………………………………….………...

**C.43

.

Which of these factors

is more important for

you to get better

health care?

Please rank your preference from 1 to 6

Waktu tempuh (Reduced travel time) ………

Waktu tunggu (Reduced waiting time) ………

Biaya (Reduced cost) ………

Faktor budaya dan agama (Cultural / religious factors) ………

Kualitas yang baik (Improved quality of service) ………

Keramahan personel (Friendliness of facility personnel) ………

C.44. Over all satisfaction

level towards existing

health care that you

are getting.



C.45. How do you think

access to health care

can be improved?

Mengurangi waktu perjalanan (Further reduced travel time)

Mengurangi waktu tunggu (Further reducing waiting time)

Mengurangi biaya (Further reduced cost)

Banyak pilihan dengan mempertimbangkan factor agama

dan tradisi (Better options regarding cultural factors)

Meningkatkan kualitas jasa pelayanan (Further improving quality

of service)

Meningkatkan personal dari pelayan rumah sakit keramahan

(Further improving personal treatment by facility personnel)

Thank you very much for your time and cooperation!!

88

Appendix D: Content of Interview with Health Facility Personnel

Sub-district: ………………….……………….

Interviewer name: ………………………..……….

Village: …………………………….………….

Position of interviewee: …………………......……

…………………………………………….……

Name of facility: …………………….………..

Type of facility:………..……………………………

Location : …..……………………………….…

(Coordinates:………………………………..…….)

Date: ……………………….…

Duration: ……………….…….

This survey is purely for study purpose to understand the existing condition in accessing health care facility in

the city of Yogyakarta, Indonesia. Individual perception of citizen is of immense value for a successful

completion of this study. Your responses to this questionnaire will be treated with strict confidentiality. Hence,

your honest comments and cooperation will be highly appreciated.

[* Read: I will ask some general questions related to the service of this health care facility]

A. General information from facility personnel

A.1. In which year did this

service start?

………………………………………………………………….

A.2. Type of service Primary examination Curative care Maternity care

Dental care Inpatient All Other

A.3. Number of medical

doctors

………………………Male

………………………Female

A.4 Number of nurse or other

medical assistants

………………………Male

………………………Female

A.5. Number of patients‟ bed ……………………… for general primary care

……………………… for emergency case

……………………… for indoor patients (admitted for few days)

A.6. Does this facility have all

laboratories and

equipments required for

primary health care?

Yes No

A.7. Does this facility provide

any extra service or

treatment equipment that

others don‟t have?

Yes No

If yes, what…………………………………………………...……..

………………………………………………………………………..

A.8. Is appointment required?

Yes No

By telephone? Yes No

If yes, before………………………………….. hours or days.

A.9. Number of patient

attendance in last month

………………………………………………… (total)

………………………………………………… ( from poor family)


89

A.10. Number of patient

attendance in last year

………………………………………………… (total)

………………………………………………… (from poor family)

A.11. Attendance of poor

patients in last 6 months

suffering from mentioned

diseases

Disease

Total

number

Number of poor patients

with card

Severe flu (cough, cold)

Fever

Diarrhea & vomiting

Malaria

Measles

Dengue fever

Upper Respiratory Tract

infection (ISPA)

Others

…………………………

A.12. Opening hour ……………………………………………………..

A.13. Does this facility provide

emergency service?

Yes No

Number of ambulance…………………… Don‟t have it

A.14. Where do majority of

patients come from?

…………………..….... Village ..………….………..Sub-district

…………………………. Regency

A.15. Cost Registration cost:……………………………………………….….

Doctor‟s fee (for primary care)n .....……………………….………

Other (if any): ……………………………………………………...

A.16. Do patients have problem

in paying?

Yes

No

A.17. Any subsidized rate of

treatment for special group

of people?

Yes

No

If yes, who……………………………………………………………

(with health card , letter of poor or others)

A.18. How far does the health

insurance cover the cost?

Cover all the cost

Cover partially (……………..……..% of total health care cost)

Does health insurance cover medication cost?

Yes No

A.19. Do patients have any

preference for male or

female doctors / medical

personnel?

Yes

No

A.20. Do patients have any

common complain?

Yes

No

If yes, what……………………………………………………….

Self observation

Amount of patients in waiting lobby…………………………………………………....…

Enough seating provision in waiting lobby…………………………………………...….…

Cleanliness ……………………………………………………………………………....……

Availability of medical store…………………………………………………………......……

General mode of transport used by patients……………………………………………..…….

Distance to nearest public transport transit …………………………………………...……

90

Appendix E: Distribution of summary scores of dimensions of

access

Figure E-1displays the distribution of summary scores for each dimension using box plot which depict

the median, quartile and outliers. The outliers are the household with values between 1.5 inter

quartile range (IQR) and 3 IQR‟s from the end value of a box. Outliers with low accessibility scores

in box plot showed in Figure (a) represented households visiting hospitals from village Tridadi.

And the outliers with higher adequacy scores were households from higher socioeconomic class who

visited private clinic from village Kricak. Figure (b) shows the distribution of summary scores for

households visiting three different health facilities. Distribution of accessibility score for hospital and

affordability score for private clinics showed wide range with low median value as compared to other

two. Adequacy score was much higher for private clinics.

Table E-1 presents the summary scores of dimensions of access for sample villages and different

healthcare facilities.


Village

Tegalpanggung

Kricak

Tridadi

0.92

0.87

0.78

0.73

0.74

0.7

0.78

0.83

0.8

0.84

0.79

0.79

0.72

0.7

0.75

Healthcare facility

Hospital

Puskesmas

Sub-puskesmas

Private clinic

0.71

0.79

0.8

0.76

0.76

0.73

0.74

0.82

0.75

0.8

0.82

0.67

0.82

0.76

0.77

0.81

0.74

0.65

0.66

0.79

Table E-1: Summary scores of dimensions of access for sample villages and different

healthcare facilities

[Note: High values indicate high level of access; 1 = maximum value]

Figure E-1: Distribution of dimensions summary scores [High score refers to higher satisfaction level; 1= highest score]

(b) Score per health facilities

1.0

0.8

0.6

0.4

0.2

0 Hospital Puskesmas Private clinic

Accessibility Availability Affordability

Acceptability Adequacy

(a) Total sample score

1.0

0.8

0.6

0.4

0.2

0
